| *** johngarbutt has quit IRC | 00:00 | |
| *** hashar has quit IRC | 00:01 | |
| *** danwent has joined #openstack-dev | 00:03 | |
| *** dubsquared has quit IRC | 00:09 | |
| *** Ryan_Lane has quit IRC | 00:12 | |
| *** PotHix has quit IRC | 00:12 | |
| *** Ryan_Lane1 has joined #openstack-dev | 00:12 | |
| *** Ryan_Lane1 is now known as Ryan_Lane | 00:13 | |
| *** Ryan_Lane has joined #openstack-dev | 00:13 | |
| *** asdfasdf has joined #openstack-dev | 00:14 | |
| *** mfer has quit IRC | 00:14 | |
| jeblair | I'm going to shut jenkins down for a gerrit-trigger rematch | 00:16 |
|---|---|---|
| *** nati has quit IRC | 00:16 | |
| *** openstackjenkins has quit IRC | 00:17 | |
| *** asdfasdf has quit IRC | 00:17 | |
| *** troytoman has quit IRC | 00:17 | |
| *** troytoman-away has joined #openstack-dev | 00:17 | |
| *** nati has joined #openstack-dev | 00:17 | |
| *** nati has joined #openstack-dev | 00:18 | |
| heckj | anotherjesse: bugs files and prioritized. Where I could (i.e. not devstack) I assigned against Essex4 | 00:19 |
| *** openstackjenkins has joined #openstack-dev | 00:19 | |
| *** heckj has quit IRC | 00:21 | |
| *** jmckenty has joined #openstack-dev | 00:21 | |
| *** adjohn has quit IRC | 00:23 | |
| *** anotherjesse has joined #openstack-dev | 00:24 | |
| jeblair | jenkins is back up | 00:25 |
| *** jmckenty has quit IRC | 00:25 | |
| *** anotherjesse has left #openstack-dev | 00:26 | |
| *** anotherjesse has joined #openstack-dev | 00:26 | |
| jdg | Anybody familiar with the use of "filter_by" in the code base? Particularly in the sqlaclchemy API? | 00:29 |
| *** littleidea has quit IRC | 00:30 | |
| andrewbogott | I have used it, although I doubt I can tell you anything that you can't learn from looking at the source... | 00:31 |
| andrewbogott | What's up? | 00:31 |
| *** littleidea has joined #openstack-dev | 00:31 | |
| jdg | andrewbogott: Well I'm familiar with filtery_by, but I don't quite understand all these seemingly always true filters, ala: filter_by(id=id) | 00:32 |
| andrewbogott | this is a guess, but... maybe one of the ids is a variable and the other refers to a field in the query? | 00:33 |
| jdg | andrewbogott: I think that's a good guess, and that was my first. But I don't know how python would tell the difference between them. | 00:34 |
| andrewbogott | Good point. | 00:35 |
| jdg | another example is (disabled=disabled), we pass in a var disabled and check against itself? | 00:35 |
| jdg | Oh wait.... | 00:35 |
| *** lloydde has quit IRC | 00:36 | |
| jdg | Maybe it's closer to what you said but just different in how it works.... | 00:36 |
| dolphm | the one on the left is not a variable, per se... it's an assignment operation, not a comparison | 00:36 |
| dolphm | filter_by(id=SOME_ID) --> WHERE `id`=SOME_ID | 00:36 |
| jdg | dolphm: right, but there is specific code where it's the same... ie filter_by(disabled=disabled) | 00:36 |
| dolphm | the one on the left isn't being evaluated | 00:37 |
| jdg | dolphm: Sorry, I'm dense. So how does the filter work? What's it's purpose? Does that mean it looks for anything in the query that has the value we've set in the filter then? | 00:38 |
| dolphm | i'm not quite sure how it works without looking it up, but I know there's a filter() and a filter_by() and the difference was non-intuitive | 00:39 |
| dolphm | but i think they both result in WHERE clauses in SQL | 00:39 |
| jdg | dolphm: Yeah, I understand the difference between them. Just can't figure out how it's working here. | 00:39 |
| jdg | column == expression vs. keyword = expression, they do the same thing really. | 00:40 |
| jdg | Ahhhh.... ok | 00:40 |
| jdg | So the expression on the left is evaluated as a key in the DB | 00:41 |
| andrewbogott | I can't for the life of me understand how python parses that. But it makes sense if I ignore that part. | 00:41 |
| *** anotherjesse1 has joined #openstack-dev | 00:42 | |
| *** anotherjesse1 has quit IRC | 00:42 | |
| jdg | So I can understand it if I assume that it means we happen to have a key in the DB that has the same name as the variable we passed in and are evaluating. | 00:42 |
| jdg | Then the DB queury is "smart" enough to know 'hey... the first item is the DB key we're looking for and we want to evaluate it against the variable on the right side of the expression. | 00:43 |
| *** jmckenty has joined #openstack-dev | 00:43 | |
| andrewbogott | I don't get why the inside of the parens isn't evaluated as an expression (True) before getting passed to filter_by. Is there some secret punctuation I'm missing? | 00:44 |
| jdg | andrewbogott: And now you and I have arrived at the same place :) | 00:44 |
| jdg | That's what I was saying to myself when I saw it, so I thought I'd shoot a question out here on IRC | 00:45 |
| andrewbogott | Maybe functions are evaluated from the outside in... | 00:45 |
| *** hub-cap has joined #openstack-dev | 00:45 | |
| *** jmckenty has quit IRC | 00:45 | |
| jdg | I think filter_by() sort of "overloads" the '=' operator to use it in a different way | 00:46 |
| jdg | One of those little pythonic things that make me beat my head against the wall the first time I see them. | 00:46 |
| andrewbogott | It's too much magic to fit in my tiny head | 00:47 |
| jdg | andrewboggot: So I'm going to spin up a devstack instance and check the database to see if this is actually what I think it is. | 00:47 |
| jdg | andrewoggot: ditto | 00:47 |
| *** cmagina has quit IRC | 00:47 | |
| jdg | Run a test against it | 00:47 |
| Kiall | jdg, actually I think thats a standard python feature (I'm not a python guy BTW) | 00:48 |
| *** cmagina has joined #openstack-dev | 00:48 | |
| Kiall | Thats named paramaters.. | 00:48 |
| Kiall | I could be wrong though ;) | 00:48 |
| andrewbogott | oh! Of course! | 00:48 |
| jdg | kiall: Yes, you're right it's standard python | 00:48 |
| andrewbogott | Kiall: That makes total sense. We were hypnotized into thinking it was an expression. | 00:49 |
| Kiall | Yea - First time I saw that was a little weird - but its actually a nice feature | 00:49 |
| *** hub_cap has quit IRC | 00:49 | |
| jdg | kiall: So it is in fact a key in the DB (or wherever you happen to be implementing the query). In these cases it just so happens we pass in the same variable name as the DB key name correct? | 00:49 |
| *** hub-cap is now known as hub_cap | 00:49 | |
| jdg | kiall: The only case I didn't understand for example were things like: filter_by(disabled=disabled) | 00:50 |
| jdg | things like "filter_by(some.thing='wtf')" I totally get. | 00:51 |
| Kiall | so `def function(first, second):...` and `function(second="B", first="A")` | 00:51 |
| Kiall | yea, disabled=disabled looks like it would evaluate to TRUE, but since its not a comparison (its closer to an assignment), it doesnt | 00:51 |
| Kiall | I think with filter_by, its a combination of named paramaters and kwargs coming into play... | 00:52 |
| jdg | kiall: See, you are a python guy :) | 00:52 |
| Kiall | I've literally done about 200 lines of python in my life ;) Learning it though! | 00:53 |
| jdg | Throwing around default assignment operations in functions and kwargs | 00:53 |
| jdg | alright, I'll bring up an install and just verify for my own curiousity that things like "disabled" are actually a key in the table | 00:54 |
| *** jakedahn has joined #openstack-dev | 00:54 | |
| *** danwent has quit IRC | 00:54 | |
| *** bencherian has joined #openstack-dev | 00:57 | |
| anotherjesse | dtroyer / termie / heckj - taking a pass at https://bugs.launchpad.net/keystone/+bug/942984 while dean is doing the devstack using service tenant/users | 00:58 |
| uvirtbot` | Launchpad bug 942984 in keystone "keystone should return 503 service unavailable when service can't get a token via admin user/pass" [Critical,Confirmed] | 00:58 |
| *** heckj has joined #openstack-dev | 01:00 | |
| heckj | ppb vote is already closed? | 01:01 |
| anotherjesse | wat | 01:01 |
| bcwaldon | reed: can you shed some light on this? | 01:02 |
| heckj | it's showing closed after 83 votes cast - guessing something broke... | 01:02 |
| reed | heckj, I emailed the mailing list with an explanation | 01:02 |
| *** nati is now known as nati_ueno | 01:02 | |
| reed | bloody launchpad is taking a f*ing nap | 01:02 |
| reed | heckj, the gist is that the ballot was missing a candidate | 01:03 |
| adam_g | keystone dudes, whats the plan for service token? is it going away, in favor of user/passwd credentials for services or something else? getting close to cutting juju charms / our CI over to the post-redux branch, but realize i can't until this is all sorted | 01:03 |
| heckj | adam_g: created a pacel of bugs related to the issue once we nailed it down. | 01:03 |
| heckj | adam_g: general description of the issues and solutions intended: http://etherpad.openstack.org/keystone-admin-config | 01:04 |
| adam_g | heckj: bodacious, thanks | 01:04 |
| *** hhoover has quit IRC | 01:05 | |
| *** dolphm has quit IRC | 01:05 | |
| heckj | adam_g: bugs all logged from that, so if you have commentary, make sure it's in the bugs. That etherpad should be just reference now | 01:05 |
| *** dolphm has joined #openstack-dev | 01:05 | |
| heckj | reed: you really should delete that election, even though I was doing great in it | 01:06 |
| *** littleidea has quit IRC | 01:06 | |
| reed | heckj, I stopped it, there is no way to delete it I think :( | 01:07 |
| *** littleidea has joined #openstack-dev | 01:08 | |
| *** dolphm has quit IRC | 01:08 | |
| *** dubsquared has joined #openstack-dev | 01:15 | |
| *** lloydde has joined #openstack-dev | 01:16 | |
| *** mnewby has quit IRC | 01:17 | |
| *** mfer has joined #openstack-dev | 01:22 | |
| *** dolphm has joined #openstack-dev | 01:22 | |
| *** sandywalsh has joined #openstack-dev | 01:23 | |
| zaitcev | What a bummer. | 01:27 |
| zaitcev | So, as it turnes out, S3 mandates that HMAC was calculated using un-decoded URL. | 01:28 |
| zaitcev | But Swift relies on WSGI, and obviously env['PATH_INFO'] is decoded. Otherwise, of course, shell scripts that were running under CGI in 1993 would not be able to open right files. | 01:30 |
| zaitcev | There is no un-decoded path available to Swift middleware. | 01:30 |
| *** Ryan_Lane is now known as Ryan_Lane|away | 01:31 | |
| zaitcev | OK, I thought, let's just hook into webob and add a special env[] for us that has un-decoded URL | 01:31 |
| *** vizsla has joined #openstack-dev | 01:31 | |
| zaitcev | I'm just wondering if creiht knew about any of this. | 01:31 |
| *** lloydde has quit IRC | 01:33 | |
| *** jdg has quit IRC | 01:34 | |
| *** apevec has quit IRC | 01:35 | |
| *** jog0 has left #openstack-dev | 01:36 | |
| *** danwent has joined #openstack-dev | 01:46 | |
| *** crobinso has quit IRC | 01:52 | |
| *** mattstep has quit IRC | 01:53 | |
| *** mattstep has joined #openstack-dev | 01:54 | |
| *** dubsquared has quit IRC | 01:54 | |
| *** adjohn has joined #openstack-dev | 01:55 | |
| *** heckj has quit IRC | 01:59 | |
| *** mfer has quit IRC | 02:02 | |
| *** x86brandon has quit IRC | 02:08 | |
| *** jog0_ has joined #openstack-dev | 02:09 | |
| *** jog0_ has quit IRC | 02:09 | |
| *** dayou has joined #openstack-dev | 02:11 | |
| *** mfer has joined #openstack-dev | 02:23 | |
| *** jakedahn has quit IRC | 02:24 | |
| *** jakedahn has joined #openstack-dev | 02:25 | |
| *** dolphm has quit IRC | 02:25 | |
| *** mattray has joined #openstack-dev | 02:26 | |
| openstackgerrit | Verification of a change to openstack/keystone failed: Provide request to Middleware.process_response() https://review.openstack.org/4616 | 02:32 |
| *** dwalleck has quit IRC | 02:32 | |
| *** eglynn_ has joined #openstack-dev | 02:32 | |
| notmyname | zaitcev: ping | 02:33 |
| zaitcev | notmyname: pong | 02:33 |
| notmyname | zaitcev: I wanted to follow up about the decoded URLs | 02:33 |
| notmyname | zaitcev: "un-decoded" == utf8? | 02:33 |
| zaitcev | No, un-decoded means "escaped" actually... e.g. https://bugs.launchpad.net/swift/+bug/936998 | 02:34 |
| uvirtbot` | Launchpad bug 936998 in swift "Object name containing colon results in 403 Forbidden" [Undecided,Confirmed] | 02:34 |
| *** dwalleck has joined #openstack-dev | 02:35 | |
| *** dwalleck has quit IRC | 02:35 | |
| *** dwalleck has joined #openstack-dev | 02:35 | |
| *** eglynn has quit IRC | 02:36 | |
| notmyname | zaitcev: I wrote a little eventlet test to look at the env | 02:36 |
| *** jdg has joined #openstack-dev | 02:36 | |
| openstackgerrit | Verification of a change to openstack/keystone failed: Align tox jobs with project standards. https://review.openstack.org/4625 | 02:37 |
| zaitcev | notmyname: I'd like to look, although from looking at eventlet/wsgi.py the situation looks pretty bleak. | 02:37 |
| notmyname | zaitcev: env['PATH_INFO'] seems to match what I pass in | 02:37 |
| zaitcev | Even if you pass %3A ? | 02:38 |
| notmyname | ah ha! | 02:39 |
| notmyname | that gets translated | 02:39 |
| reed | new ballot for PPB election is out | 02:39 |
| reed | and I go to have dinner | 02:39 |
| notmyname | reed: is the first one bad? I need to vote again? | 02:39 |
| zaitcev | it happens here, as far as I understand: https://bitbucket.org/which_linden/eventlet/src/a61f71411e83/eventlet/wsgi.py | 02:39 |
| notmyname | reed: sorry, just saw your email | 02:39 |
| zaitcev | class HttpProtocol(BaseHTTPServer.BaseHTTPRequestHandler): | 02:40 |
| reed | notmyname, the first one will tell you that the poll is closed | 02:40 |
| reed | the new one... I hope it has enough details so that people will figure out what's happening before they clog my email | 02:40 |
| zaitcev | handler's self has self.path which is not decoded, but I do not see any way at all to get at it from the outside. | 02:41 |
| reed | it was going so well so far... I brought my inbox down to 5 unread/unacted messages and I guess tomorrow it will have 50 and more again | 02:41 |
| reed | oh well | 02:41 |
| reed | bye folks, I go get drunk with other Italians | 02:41 |
| *** reed has quit IRC | 02:41 | |
| termie | mtaylor: how do i go about removing people from core in keystone, btw? | 02:42 |
| *** adjohn has quit IRC | 02:42 | |
| anotherjesse | mtaylor: why did your change break pep8 - https://review.openstack.org/#change,4625 | 02:43 |
| termie | anotherjesse: all changes are triggering pep8 problems | 02:44 |
| anotherjesse | k | 02:44 |
| termie | anotherjesse: i still don't know why, but anyway, other changes hit the same thign | 02:45 |
| anotherjesse | going to play with my daughter while someone fixes ;) when she sleeps I hope to fix authtoken | 02:45 |
| notmyname | zaitcev: seems like this would be an issue with eventlet | 02:45 |
| zaitcev | notmyname: Indeed. BTW, there's no urgency in fixing it. I don't have a customer hitting it. I just thought it would be a perfect beginner's bug to fix... I expected that someone forgot a split(':') or had one too many somewhere. | 02:46 |
| *** jakedahn has quit IRC | 02:46 | |
| zaitcev | brb tea | 02:46 |
| *** jakedahn has joined #openstack-dev | 02:47 | |
| notmyname | zaitcev: ah, I see dweimer was asking about it in #eventlet earlier | 02:48 |
| *** shevek__ has quit IRC | 02:49 | |
| *** dolphm has joined #openstack-dev | 02:50 | |
| *** jog0 has joined #openstack-dev | 02:50 | |
| *** jdurgin has quit IRC | 02:53 | |
| *** jog0 has left #openstack-dev | 02:56 | |
| *** jdurgin has joined #openstack-dev | 02:58 | |
| *** andrewbogott is now known as andrewbogott_afk | 03:00 | |
| *** jakedahn has quit IRC | 03:04 | |
| *** jakedahn___ has joined #openstack-dev | 03:05 | |
| *** jakedahn has joined #openstack-dev | 03:06 | |
| *** jakedahn has quit IRC | 03:07 | |
| *** jakedahn___ has quit IRC | 03:07 | |
| *** jakedahn has joined #openstack-dev | 03:07 | |
| *** novas0x2a|laptop has quit IRC | 03:10 | |
| *** dolphm has quit IRC | 03:14 | |
| *** bengrue has quit IRC | 03:16 | |
| *** dolphm has joined #openstack-dev | 03:19 | |
| *** mdomsch has joined #openstack-dev | 03:20 | |
| *** zul has quit IRC | 03:23 | |
| notmyname | anotherjesse: ping | 03:23 |
| notmyname | anotherjesse: nm | 03:23 |
| notmyname | vishy: ping | 03:23 |
| anotherjesse | notmyname: ping | 03:24 |
| anotherjesse | notmyname: nm | 03:25 |
| notmyname | anotherjesse: sorry, I was trying to remember how to do something with gerrit that vishy and I talked about | 03:25 |
| notmyname | I've got a good patch proposed, but it has one small pep8 issue | 03:27 |
| notmyname | I'd prefer not to have to send it back. we had talked about how to make the small patch, but I don't remember the details | 03:27 |
| *** littleidea has quit IRC | 03:29 | |
| *** HugoKuo_ has quit IRC | 03:32 | |
| *** vizsla has quit IRC | 03:35 | |
| openstackgerrit | Verification of a change to openstack/quantum failed: remove pep8 and strict lxml version from setup.py https://review.openstack.org/4235 | 03:36 |
| *** littleidea has joined #openstack-dev | 03:37 | |
| *** mdomsch has quit IRC | 03:40 | |
| *** dtroyer has quit IRC | 03:41 | |
| *** jog0 has joined #openstack-dev | 03:46 | |
| *** maplebed has quit IRC | 03:46 | |
| *** x86brandon has joined #openstack-dev | 03:53 | |
| *** mfer has quit IRC | 03:54 | |
| *** x86brandon_ has joined #openstack-dev | 03:54 | |
| anotherjesse | notmyname - as core, you can check the branch out, make some changes, git add changes, git commit —amend, then git review updates it | 03:54 |
| anotherjesse | (to anyone's branch) | 03:54 |
| anotherjesse | they retain authorship (I think) | 03:54 |
| notmyname | ah | 03:56 |
| notmyname | thanks | 03:56 |
| *** jog0 has quit IRC | 03:56 | |
| notmyname | that's with git review -d <patch number>? | 03:56 |
| anotherjesse | notmyname: perhaps - I just use the gerrit instructions for checking out the patchset | 03:57 |
| *** davlap has quit IRC | 03:57 | |
| *** vizsla has joined #openstack-dev | 03:57 | |
| notmyname | ok. I'll look at the docs some more. I've only used git review to submit stuff, not to download | 03:57 |
| anotherjesse | I know it works if you do the checkout section | 03:58 |
| *** x86brandon has quit IRC | 03:58 | |
| *** x86brandon_ is now known as x86brandon | 03:58 | |
| *** Mandell has quit IRC | 04:01 | |
| *** adjohn has joined #openstack-dev | 04:12 | |
| dweimer | notmyname: I came to the same conclusion as zaitcev. The only workaround I've found so far is to set a separate env variable in eventlet:wsgi.py with self.path and then use that in swift3. | 04:13 |
| *** Mandell has joined #openstack-dev | 04:13 | |
| notmyname | dweimer: which doesn't actually sound like too bad an idea. normally, you want what path_info gives you. but not having any access at all to what what "on the wire" seems strange | 04:14 |
| openstackgerrit | Verification of a change to openstack/horizon failed: Adds usage vs quota data to the launch instance dialog. Adds a reusable progress bar indicator. https://review.openstack.org/4670 | 04:15 |
| notmyname | dweimer: if you've not submitted code to eventlet, they are pretty easy to deal with, from what I hear. we (the swift team, and I think other openstack teams too) have submitted several patches that have been accepted | 04:15 |
| *** bencherian has quit IRC | 04:15 | |
| *** bencherian has joined #openstack-dev | 04:15 | |
| dweimer | notmyname: That's good to hear. I'll add a test for my change and send it in. Do they have any external review process like gerrit or is it just a pull request through bitbucket? | 04:17 |
| *** dwalleck has quit IRC | 04:17 | |
| notmyname | dweimer: no, all through bitbucket, I think | 04:18 |
| dweimer | Regarding swift3, should it be using the decoded path_info when writing objects to the object server? Right now it replaces path_info with a quoted version so you get objects like foo%40bar sent to the object server. | 04:25 |
| openstackgerrit | Verification of a change to openstack/quantum failed: remove pep8 and strict lxml version from setup.py https://review.openstack.org/4235 | 04:25 |
| *** jdg has quit IRC | 04:30 | |
| *** mattray has quit IRC | 04:33 | |
| zaitcev | dweimer: I think it should, so that after the fix the objects remain where they were - in case of both S3 and naive (obviously you could not write them through S3 before). | 04:45 |
| *** maplebed has joined #openstack-dev | 04:47 | |
| *** Gordonz has joined #openstack-dev | 04:47 | |
| *** hub_cap has quit IRC | 04:49 | |
| dweimer | zaitcev: I gree. I'll test against S3 first to make sure that's what Amazon does, but I wouldn't expect them to have container listings full of encoded characters. A patch will have to wait until there is a fix for the authentication issue though. If we save the decoded version now, swift3 users can't get the file back. | 04:52 |
| *** anotherjesse1 has joined #openstack-dev | 04:53 | |
| *** Ryan_Lane|away has quit IRC | 04:55 | |
| *** vizsla has quit IRC | 05:00 | |
| *** jmckenty has joined #openstack-dev | 05:05 | |
| *** troytoman-away has quit IRC | 05:06 | |
| *** troytoman-away has joined #openstack-dev | 05:06 | |
| openstackgerrit | Verification of a change to openstack/horizon failed: Implements reusable tab components. https://review.openstack.org/4655 | 05:06 |
| *** lloydde has joined #openstack-dev | 05:10 | |
| *** jmckenty has quit IRC | 05:19 | |
| openstackgerrit | Verification of a change to openstack/nova failed: Return empty list when volume not attached https://review.openstack.org/4663 | 05:29 |
| *** Gordonz has quit IRC | 05:39 | |
| *** hattwick has quit IRC | 05:41 | |
| *** adjohn has quit IRC | 05:59 | |
| *** adjohn has joined #openstack-dev | 06:02 | |
| *** adjohn has quit IRC | 06:12 | |
| *** Ryan_Lane has joined #openstack-dev | 06:16 | |
| *** tryggvil_ has quit IRC | 06:17 | |
| *** tryggvil_ has joined #openstack-dev | 06:20 | |
| *** bepernoot has joined #openstack-dev | 06:34 | |
| *** danwent has quit IRC | 06:38 | |
| *** hattwick has joined #openstack-dev | 06:38 | |
| *** maplebed_ has joined #openstack-dev | 06:47 | |
| *** maplebed_ has joined #openstack-dev | 06:48 | |
| *** maplebed is now known as Guest56872 | 06:48 | |
| *** maplebed_ is now known as maplebed | 06:48 | |
| *** Guest56872 has quit IRC | 06:50 | |
| *** nati_ueno has quit IRC | 06:56 | |
| *** bepernoot has quit IRC | 06:57 | |
| *** x86brandon has quit IRC | 06:59 | |
| vishy | notmyname: don't know if you figured it out but git review -d XXXX will create the branch for you | 07:05 |
| *** nati_ueno has joined #openstack-dev | 07:05 | |
| vishy | then you can fiddle, git commit --amend -a and git review | 07:05 |
| *** nati_ueno has quit IRC | 07:06 | |
| *** berendt has joined #openstack-dev | 07:10 | |
| *** shevek__ has joined #openstack-dev | 07:18 | |
| *** berendt has quit IRC | 07:20 | |
| *** gkotton has joined #openstack-dev | 07:21 | |
| *** gkotton has quit IRC | 07:24 | |
| *** gkotton has joined #openstack-dev | 07:25 | |
| LarsErikP | where does dashboard log? | 07:34 |
| anotherjesse1 | apache log | 07:34 |
| *** berendt has joined #openstack-dev | 07:35 | |
| LarsErikP | by default? | 07:35 |
| *** berendt has quit IRC | 07:35 | |
| anotherjesse1 | it is a normal django app - so wherever you normally put it | 07:36 |
| *** gkotton has joined #openstack-dev | 07:36 | |
| *** hattwick has quit IRC | 07:41 | |
| *** dalang has quit IRC | 07:48 | |
| *** hattwick has joined #openstack-dev | 07:53 | |
| *** bencherian has quit IRC | 07:58 | |
| *** bencherian has joined #openstack-dev | 07:59 | |
| *** bencherian has quit IRC | 08:03 | |
| LarsErikP | how can this be reported as an error? SELECT `django_session`.`session_key`, `django_session`.`session_data`, `django_session`.`expire_date` FROM `django_session` WHERE (`django_session`.`session_key` = 6b2600168283e84ba110e625050f314b AND `django_session`.`expire_date` > 2012-02-29 08:51:01 ); | 08:05 |
| anotherjesse1 | looks like you haven't sync'd the db | 08:06 |
| LarsErikP | aha | 08:06 |
| LarsErikP | there we have an interesting problem.. because.. wait for it | 08:06 |
| LarsErikP | root@dublin:~# nova-manage db sync | 08:06 |
| LarsErikP | Command failed, please check log for more info | 08:06 |
| LarsErikP | root@dublin:~# tail /var/log/nova/nova-manage.log | 08:07 |
| LarsErikP | (nova): TRACE: File "/usr/lib/python2.7/dist-packages/migrate/versioning/schema.py", line 80, in changeset | 08:07 |
| LarsErikP | (nova): TRACE: changeset = self.repository.changeset(database, start_ver, version) | 08:07 |
| LarsErikP | (nova): TRACE: File "/usr/lib/python2.7/dist-packages/migrate/versioning/repository.py", line 214, in changeset | 08:07 |
| LarsErikP | (nova): TRACE: changes = [self.version(v).script(database, op) for v in versions] | 08:07 |
| LarsErikP | (nova): TRACE: File "/usr/lib/python2.7/dist-packages/migrate/versioning/repository.py", line 178, in version | 08:07 |
| LarsErikP | (nova): TRACE: return self.versions.version(*p, **k) | 08:07 |
| LarsErikP | (nova): TRACE: File "/usr/lib/python2.7/dist-packages/migrate/versioning/version.py", line 125, in version | 08:07 |
| LarsErikP | (nova): TRACE: return self.versions[VerNum(vernum)] | 08:07 |
| LarsErikP | (nova): TRACE: KeyError: <VerNum(74)> | 08:07 |
| LarsErikP | (nova): TRACE: | 08:07 |
| LarsErikP | any ideas? | 08:13 |
| *** dolphm has quit IRC | 08:13 | |
| anotherjesse1 | LarsErikP: have you checked out what stack.sh does in https://github.com/openstack-dev/devstack/tree/stable/diablo to setup dashboard? | 08:15 |
| *** adjohn has joined #openstack-dev | 08:17 | |
| comstud | LarsErikP: looks like maybe migration 74 is missing, but you have a 75 somehow | 08:18 |
| *** jakedahn__ has joined #openstack-dev | 08:18 | |
| LarsErikP | actually, we used the managedIT scripts for initial install. I think I screwed it up, when i ran nova-manage db sync from my compute nodes.. After i did that, the command failed on my controller... | 08:18 |
| LarsErikP | so, what I'm trying to say... it worked before i did that | 08:19 |
| LarsErikP | db version on the controller yields 74 | 08:19 |
| *** jakedahn has quit IRC | 08:22 | |
| *** jakedahn__ is now known as jakedahn | 08:22 | |
| *** hashar has joined #openstack-dev | 08:30 | |
| *** littleidea has quit IRC | 08:37 | |
| *** tomoe_ has quit IRC | 08:42 | |
| LarsErikP | ?? | 08:47 |
| *** adjohn has quit IRC | 08:54 | |
| *** sandywalsh has quit IRC | 09:03 | |
| *** anotherjesse1 has quit IRC | 09:03 | |
| *** Ryan_Lane has quit IRC | 09:08 | |
| *** darraghb has joined #openstack-dev | 09:09 | |
| *** dneary has joined #openstack-dev | 09:09 | |
| *** dneary has quit IRC | 09:09 | |
| *** dneary has joined #openstack-dev | 09:09 | |
| *** berendt has joined #openstack-dev | 09:12 | |
| *** sandywalsh has joined #openstack-dev | 09:16 | |
| *** derekh has joined #openstack-dev | 09:20 | |
| *** zigo has joined #openstack-dev | 09:24 | |
| *** gkotton has quit IRC | 09:24 | |
| *** zigo has quit IRC | 09:29 | |
| *** Mandell has quit IRC | 09:31 | |
| *** anotherjesse1 has joined #openstack-dev | 09:35 | |
| *** zigo has joined #openstack-dev | 09:35 | |
| *** anotherjesse1 has quit IRC | 09:42 | |
| *** yamahata__ has quit IRC | 09:43 | |
| *** yamahata_ has quit IRC | 09:43 | |
| *** yamahata has joined #openstack-dev | 09:46 | |
| *** gooberguy has joined #openstack-dev | 09:47 | |
| *** popux has joined #openstack-dev | 09:59 | |
| *** pixelbeat has joined #openstack-dev | 10:03 | |
| *** apevec has joined #openstack-dev | 10:24 | |
| *** bepernoot has joined #openstack-dev | 10:32 | |
| *** mikemowgli has quit IRC | 10:37 | |
| *** Ryan_Lane has joined #openstack-dev | 10:38 | |
| *** derekh has quit IRC | 10:39 | |
| *** paulormg has joined #openstack-dev | 10:42 | |
| *** Ryan_Lane has quit IRC | 10:48 | |
| *** derekh has joined #openstack-dev | 10:54 | |
| *** berendt has quit IRC | 10:54 | |
| *** zigo has quit IRC | 10:58 | |
| *** Mkenneth has joined #openstack-dev | 11:07 | |
| *** zul has joined #openstack-dev | 11:16 | |
| *** rods has joined #openstack-dev | 11:33 | |
| *** oneiroi has joined #openstack-dev | 11:35 | |
| *** yamahata__ has joined #openstack-dev | 11:41 | |
| *** Lxu has joined #openstack-dev | 11:47 | |
| LarsErikP | uhm.. now nova-network is rude | 11:49 |
| LarsErikP | trying to fetch columns from tables that doesn't exist | 11:49 |
| LarsErikP | OperationalError: (OperationalError) (1054, "Unknown column 'instances_1.local_gb' in 'field list'") | 11:51 |
| *** journeeman has joined #openstack-dev | 12:06 | |
| *** sandywalsh has quit IRC | 12:07 | |
| *** maploin has joined #openstack-dev | 12:10 | |
| *** maploin has quit IRC | 12:10 | |
| *** maploin has joined #openstack-dev | 12:10 | |
| *** bepernoot has quit IRC | 12:10 | |
| *** hashar has quit IRC | 12:16 | |
| *** dtroyer has joined #openstack-dev | 12:19 | |
| *** sandywalsh has joined #openstack-dev | 12:20 | |
| eglynn_ | any glance core reviewers on-line right now? | 12:28 |
| *** vizsla has joined #openstack-dev | 12:32 | |
| *** dneary has quit IRC | 12:33 | |
| *** dneary has joined #openstack-dev | 12:41 | |
| *** dneary has quit IRC | 12:41 | |
| *** dneary has joined #openstack-dev | 12:41 | |
| LarsErikP | fixed nova-network.. | 12:43 |
| *** markvoelker has joined #openstack-dev | 12:44 | |
| *** journeeman has quit IRC | 12:50 | |
| LarsErikP | anotherjesse: wich db, btw? | 12:50 |
| *** dprince has joined #openstack-dev | 13:03 | |
| *** gkotton has joined #openstack-dev | 13:03 | |
| *** gkotton has quit IRC | 13:03 | |
| *** gakott has joined #openstack-dev | 13:03 | |
| *** andrewsmedina has quit IRC | 13:03 | |
| *** ches has quit IRC | 13:06 | |
| *** hashar has joined #openstack-dev | 13:06 | |
| *** ches has joined #openstack-dev | 13:06 | |
| *** egallen has joined #openstack-dev | 13:08 | |
| Daviey | Top tip for today, make sure the tests pass before you start hacking... otherwise you spent too much time trying to work out what you did to break it :) | 13:09 |
| Daviey | fml | 13:09 |
| *** gakott has quit IRC | 13:10 | |
| *** stuntmachine has joined #openstack-dev | 13:11 | |
| *** gkotton has joined #openstack-dev | 13:12 | |
| Kiall | Daviey: lol.. done that before ;) | 13:12 |
| *** pixelbeat has quit IRC | 13:16 | |
| maploin | how can I retrigger a failed Jenkins build (its fault, not the commiter's) from Gerrit? | 13:25 |
| Daviey | maploin: Mark it approved again? | 13:26 |
| maploin | Daviey: what if I'm the committer, not the reviewer? | 13:27 |
| Daviey | maploin: you can retrigger from within jenkins. | 13:27 |
| Daviey | oh, sad out of luck. | 13:27 |
| maploin | :) | 13:27 |
| maploin | I can't retrigger from Jenkins either if I'm the committer? | 13:27 |
| *** andrewsmedina has joined #openstack-dev | 13:29 | |
| Daviey | maploin: no, pester a core dev of the component you are working on | 13:31 |
| maploin | ok, thanks | 13:31 |
| *** rbasak has joined #openstack-dev | 13:31 | |
| *** mfer has joined #openstack-dev | 13:33 | |
| *** mfer has quit IRC | 13:34 | |
| *** PotHix has joined #openstack-dev | 13:36 | |
| *** ayoung has joined #openstack-dev | 13:37 | |
| *** pixelbeat has joined #openstack-dev | 13:42 | |
| *** gkotton has quit IRC | 13:42 | |
| *** bsza has joined #openstack-dev | 13:43 | |
| *** eglynn__ has joined #openstack-dev | 13:43 | |
| LarsErikP | it | 13:44 |
| LarsErikP | wrong shell | 13:44 |
| *** eglynn_ has quit IRC | 13:46 | |
| *** littleidea has joined #openstack-dev | 13:49 | |
| openstackgerrit | Verification of a change to openstack/glance failed: Disallow file:// sources on location or copy-from. https://review.openstack.org/4602 | 13:54 |
| *** crobinso has joined #openstack-dev | 13:57 | |
| PotHix | Someone working with quantum-client? | 13:57 |
| blamar | PotHix: worked with it a little, whats up? | 13:59 |
| PotHix | blamar: How are you running the unit tests? | 14:00 |
| blamar | PotHix: using 'tox' | 14:01 |
| blamar | only running python 2.6 environment though | 14:01 |
| blamar | i think it's 'tox -e py26' | 14:01 |
| PotHix | I'm using Python 2.7.2 for quantum, the 2.6 returns some errors for me | 14:03 |
| PotHix | blamar: Do you have other option to suggest? | 14:03 |
| blamar | PotHix: Can you put the errors on paste.openstack.org? I might be able to assist | 14:03 |
| blamar | How are you running the tests? | 14:04 |
| PotHix | These errors I got on the quantum repository | 14:04 |
| PotHix | I don't know how to run the tests for quantum client | 14:04 |
| PotHix | can you help me with that? | 14:05 |
| *** zykes- has joined #openstack-dev | 14:05 | |
| blamar | PotHix: potentially I can, lemme test something really quick on a python 2.7 machine | 14:06 |
| *** Mkenneth has quit IRC | 14:06 | |
| *** hub_cap has joined #openstack-dev | 14:07 | |
| *** hub_cap has quit IRC | 14:07 | |
| *** Mkenneth has joined #openstack-dev | 14:07 | |
| PotHix | blamar: which command are you using to run the unit tests? | 14:08 |
| blamar | PotHix: 'tox -e py27' just worked for me on ubuntu | 14:08 |
| PotHix | blamar: I'll try it! :) | 14:11 |
| *** lts has joined #openstack-dev | 14:14 | |
| PotHix | it works! Tks blamar! | 14:15 |
| *** pixelbeat_ has joined #openstack-dev | 14:16 | |
| *** gkotton has joined #openstack-dev | 14:16 | |
| *** derekh has quit IRC | 14:23 | |
| *** rkukura has quit IRC | 14:31 | |
| *** flaviamissi has joined #openstack-dev | 14:32 | |
| mjfork | any one else having problems with horizon either via github or running tests with a new venv? | 14:35 |
| *** derekh has joined #openstack-dev | 14:37 | |
| *** pixelbeat has quit IRC | 14:42 | |
| *** sandywalsh has quit IRC | 14:42 | |
| *** mattray has joined #openstack-dev | 14:47 | |
| *** joesavak has joined #openstack-dev | 14:48 | |
| *** mdomsch has joined #openstack-dev | 14:49 | |
| *** mattray has quit IRC | 14:49 | |
| *** mattray has joined #openstack-dev | 14:50 | |
| *** hashar has quit IRC | 14:55 | |
| *** mdomsch has quit IRC | 14:56 | |
| eglynn__ | nova folks: quick question | 14:57 |
| *** sandywalsh has joined #openstack-dev | 14:57 | |
| eglynn__ | I'm seeing odd transient failures in the Jenkins devstack integ tests gating a glance merge | 14:58 |
| eglynn__ | looks like an instance isn't starting up quickly enough | 14:58 |
| eglynn__ | timeout 60 sh -c 'while ! nova show 8b4b3067-a7d2-4fd7-b8d1-b4069660d6b2 | grep status | grep -q ACTIVE; do sleep 1; done' | 14:58 |
| eglynn__ | server didn't become active! | 14:58 |
| eglynn__ | + echo 'server didn'\''t become active!' | 14:58 |
| eglynn__ | ^^^ that failure ring any bells with anyone? | 14:59 |
| uvirtbot` | eglynn__: Error: "^^" is not a valid command. | 14:59 |
| eglynn__ | here's the full context if anyone is interested ... https://jenkins.openstack.org/job/gate-integration-tests-devstack-vm/1941/console | 15:00 |
| *** pixelbeat_ has quit IRC | 15:01 | |
| *** pixelbeat has joined #openstack-dev | 15:02 | |
| *** eglynn__ has quit IRC | 15:15 | |
| *** pmyers has quit IRC | 15:17 | |
| *** rkukura has joined #openstack-dev | 15:17 | |
| *** pmyers has joined #openstack-dev | 15:17 | |
| *** pmyers has joined #openstack-dev | 15:19 | |
| annegentle | ttx: around? I have a Q about what to do with bugs once they are marked "Fix released" but the are not targeted to a milestone (doc bugs are like this). Do I have to target to a milestone to close a bug? | 15:24 |
| *** littleidea has quit IRC | 15:24 | |
| *** jsavak has joined #openstack-dev | 15:25 | |
| ttx | annegentle: openstack-manuals bugs ? Or project bugs ? | 15:25 |
| *** nati_ueno has joined #openstack-dev | 15:26 | |
| ttx | the idea of adding the milestone is to be able to look at the milestone page and ~see what it fixed | 15:26 |
| annegentle | ttx: specifically openstack-manuals bugs | 15:26 |
| ttx | annegentle: since openstack-manuals is not formally released at each milestone... I don't see the point | 15:26 |
| annegentle | ttx: me neither. all I want to do is close bugs when I know they are fixed. | 15:26 |
| ttx | annegentle: then don't go through the hassle of targeting post-fix | 15:27 |
| annegentle | ttx: published to trunk is "Fix Released" for openstack-manuals (although another definition could be "published to release name" | 15:27 |
| ttx | annegentle: sounds fine to me | 15:28 |
| annegentle | ttx: I'm still missing a step though. What marks a bug "closed" | 15:28 |
| ttx | annegentle: FixReleased, Invalid, Opinion, Wontfix | 15:28 |
| *** joesavak has quit IRC | 15:28 | |
| ttx | annegentle: all those won't apear by default in bug searches | 15:29 |
| ttx | appear* | 15:29 |
| annegentle | ttx: ok, and this may be a CI question, but if I know a branch was merged yesterday, should the bug be updated with a link to the Gerrit branch? | 15:29 |
| annegentle | mtaylor: or jeblair ^^ trying to see if openstack-manuals is behaving as it should be connected to Gerrit. | 15:29 |
| ttx | annegentle: that's useful info. Gerrit auto-updates the bug when it has the info | 15:30 |
| annegentle | for example, I fixed 940416 yesterday. I don't see an auto-update. | 15:30 |
| ttx | annegentle: I wrote the update_bug script, let me check | 15:30 |
| annegentle | ttx: awesome, let's figure this out | 15:30 |
| ttx | annegentle: what was the corresponding review ? | 15:31 |
| annegentle | ttx: ah, never mind! It didn't actually get reviewed yet. https://review.openstack.org/#change,4652 | 15:32 |
| annegentle | ttx: my mistake | 15:32 |
| ttx | haha | 15:32 |
| ttx | annegentle: though it won't work when it does | 15:32 |
| annegentle | ttx: here's one with the Gerrit info https://bugs.launchpad.net/openstack-manuals/+bug/882781, but I have to manually mark it "Fix Released" yes? | 15:32 |
| uvirtbot` | Launchpad bug 882781 in openstack-manuals "Flag for multi_host flag is missing from docs" [Medium,Fix committed] | 15:32 |
| ttx | you need to say "bug 940416" | 15:32 |
| uvirtbot` | Launchpad bug 940416 in openstack-manuals "On api.openstack.org site, image API doesn't have a version number like v1 similar to the other APIs" [High,Fix released] https://launchpad.net/bugs/940416 | 15:32 |
| ttx | not just "940416" | 15:32 |
| ttx | in the commit message | 15:33 |
| annegentle | ttx: ah ha! | 15:33 |
| ttx | then it will link to it when you propose... | 15:33 |
| ttx | and mark it fixcommitted when it merges | 15:33 |
| *** Gordonz has joined #openstack-dev | 15:33 | |
| ttx | though in your case you'd prefer if it was directly marked FixReleased. | 15:33 |
| ttx | annegentle: file an openstack-ci bug if you want update_bug to directly set merged changes to FixReleased for openstack-manuals | 15:34 |
| annegentle | ttx: ok, that's fixed | 15:34 |
| annegentle | ttx: ah, that might be what I want. | 15:34 |
| *** Gordonz has quit IRC | 15:34 | |
| annegentle | ttx: tell me if you think that's a bad idea though. | 15:34 |
| notmyname | mtaylor: I've got a swift-core member that isn't showing up as core in gerrit (jay payne, aka letterj) | 15:34 |
| ttx | no, makes sense. | 15:35 |
| annegentle | ttx: ok, thanks | 15:35 |
| *** Gordonz has joined #openstack-dev | 15:35 | |
| *** dubsquared has joined #openstack-dev | 15:36 | |
| *** mdomsch has joined #openstack-dev | 15:39 | |
| gkotton | hi, anyone have any difficulties with devstack at the moment? | 15:40 |
| *** Mkenneth has quit IRC | 15:40 | |
| *** Mkenneth has joined #openstack-dev | 15:41 | |
| *** eglynn has joined #openstack-dev | 15:41 | |
| *** nati_ueno has quit IRC | 15:41 | |
| *** nati_ueno has joined #openstack-dev | 15:42 | |
| *** rkukura has left #openstack-dev | 15:42 | |
| *** hashar has joined #openstack-dev | 15:43 | |
| ironcamel | gkotton: not at the moment. what issue are you having? | 15:45 |
| *** rkukura has joined #openstack-dev | 15:45 | |
| gkotton | ironcamel: ./stack.sh: line 619: cd: /opt/stack/nova: No such file or directory | 15:46 |
| *** cp16net has joined #openstack-dev | 15:46 | |
| ironcamel | ummm, do you have a /opt/stack/nova dir? | 15:46 |
| ironcamel | stack.sh actually should create that itself | 15:47 |
| ironcamel | i wonder if it is a file system permissions problem | 15:47 |
| ironcamel | what user are you running as? root, stack, another user? | 15:48 |
| gkotton | ironcamel: no, the directory does not exist - this is odd! | 15:48 |
| jk0 | would be cool if LP had some sort of cap on all of these bug update emails | 15:48 |
| *** dolphm has joined #openstack-dev | 15:48 | |
| *** eglynn has quit IRC | 15:50 | |
| *** kbringard has joined #openstack-dev | 15:50 | |
| *** eglynn has joined #openstack-dev | 15:53 | |
| *** nati_ueno has quit IRC | 15:54 | |
| eglynn | bcwaldon: did my response on https://review.openstack.org/#change,4645 address your concerns on the rate limiting duplication? | 15:55 |
| milner | Any trick to get jenkins to test my approved branch? https://review.openstack.org/#change,4459 | 15:57 |
| mtaylor | notmyname: k. will figure out why | 15:58 |
| *** littleidea has joined #openstack-dev | 15:58 | |
| notmyname | mtaylor: thanks | 15:58 |
| mtaylor | actually... | 15:59 |
| mtaylor | LinuxJedi: ^^^ | 15:59 |
| mtaylor | LinuxJedi: notmyname reports that letterj/Jay Payne is in the swift-core team but not showing up in gerrit | 15:59 |
| LinuxJedi | mtaylor: ok, I shall look into it shortly | 16:00 |
| *** hhoover has joined #openstack-dev | 16:00 | |
| *** danwent has joined #openstack-dev | 16:02 | |
| zul | mtaylor jeblair: the horizon tarballs is failing (we use it to build debs) can you see why? | 16:05 |
| eglynn | jpipes: can you re-trigger Jenkins build for https://review.openstack.org/#change,4602 ? | 16:06 |
| *** mattray has quit IRC | 16:06 | |
| eglynn | jpipes: unreproducable glance test failure didn't occur in the last 2 runs, but now seeing an intermittent devstack failure seemingly caused by slow nova instance start-up | 16:07 |
| *** dubsquared has quit IRC | 16:08 | |
| *** joesavak has joined #openstack-dev | 16:08 | |
| *** jsavak has quit IRC | 16:11 | |
| *** egallen has quit IRC | 16:12 | |
| LinuxJedi | mtaylor, notmyname: gerrit is showing Jay Payne as a swift-core member for me. Maybe the sync script hadn't caught up at the time? | 16:15 |
| *** aweiss has joined #openstack-dev | 16:15 | |
| *** gkotton has quit IRC | 16:15 | |
| notmyname | LinuxJedi: ya, he's listed in core in launchpad, but he can't do +-2 reviews or merge things | 16:15 |
| LinuxJedi | notmyname: he should be able to, gerrit says he can | 16:16 |
| zaitcev | dweimer: Thanks for the update, I was just going to poke eventlet ppl this morning... | 16:18 |
| LinuxJedi | notmyname: unfortunately the current version of the script doesn't log when he was added to it in gerrit (the new version coming soon will), so I can't say when that happened | 16:18 |
| notmyname | LinuxJedi: actually, he should have always been a member of core | 16:18 |
| *** gkotton has joined #openstack-dev | 16:20 | |
| LinuxJedi | notmyname: ah, I see the problem, he is coming in via. a different launchpad OpenID | 16:20 |
| notmyname | LinuxJedi: ah. can the mapping be changed to the one he's using? | 16:21 |
| notmyname | LinuxJedi: I don't think he intends to have 2 ids | 16:21 |
| LinuxJedi | notmyname: it depends why that is happening. I'll look into it now | 16:21 |
| *** andrewsben has joined #openstack-dev | 16:24 | |
| LinuxJedi | notmyname: has he had 2 LaunchPad account merged (or does he have 2 active accounts even?) | 16:25 |
| notmyname | LinuxJedi: I doubt he has two active accounts | 16:27 |
| *** dneary has quit IRC | 16:28 | |
| *** zzed has joined #openstack-dev | 16:29 | |
| *** heckj has joined #openstack-dev | 16:31 | |
| *** andrewbogott_afk is now known as andrewbogott | 16:32 | |
| LinuxJedi | notmyname: all I can see so far is yesterday he started coming in via. a different openID, almost as if he was logged into someone else's LP account. I really need to know if he has had accounts merged at any time and if he is really logged in as him in LP. | 16:33 |
| notmyname | LinuxJedi: I'll ask | 16:34 |
| mtaylor | zul: it's because horizon got fixed to behave like normal! | 16:34 |
| * mtaylor kisses horizon on the mouth | 16:35 | |
| mtaylor | zul: it's going to be a little while before I can fix that - but I'm on it | 16:35 |
| LinuxJedi | mtaylor: you don't know where that has been ;) | 16:35 |
| notmyname | LinuxJedi: yes, he had 2 LP accounts, and they were merged | 16:35 |
| notmyname | LinuxJedi: he had a @rackspace account and a @gmail account. he moved everything to the @gmail one | 16:36 |
| *** Vek has quit IRC | 16:37 | |
| LinuxJedi | notmyname: fantastic. It is a Launchpad bug. We have instructions somewhere on what he needs to do. I'll dig them out | 16:37 |
| LinuxJedi | notmyname: ok, full instructions (with his openIDs) are here: http://paste.openstack.org/show/5472/ | 16:39 |
| notmyname | LinuxJedi: thanks | 16:40 |
| mtaylor | notmyname: we're hoping that if we file enough of these, we'll convince the person who thinks it's not a bug that it's a bug | 16:40 |
| mtaylor | :) | 16:40 |
| notmyname | mtaylor: happy to help :-) | 16:40 |
| LinuxJedi | mtaylor: the good news is now that it has started to break Launchpad projects they do consider it a bug | 16:40 |
| mtaylor | LinuxJedi: we should investigate whether or not adding a second openid entry in the gerrit db is a valid workaround | 16:40 |
| mtaylor | LinuxJedi: excellent | 16:41 |
| mtaylor | LinuxJedi: they should listen to me more quickly | 16:41 |
| *** dolphm has quit IRC | 16:41 | |
| LinuxJedi | mtaylor: it is a workaround but the sync script won't know about it if we have to rebuild the DB at any time | 16:41 |
| LinuxJedi | mtaylor: and I don't know any workaround for that :) | 16:41 |
| LinuxJedi | mtaylor: will, they have like a gazillion other bugs in the openID system so it is just when they find time :) | 16:42 |
| jeblair | mtaylor: it would also require significant identity verification on our part | 16:42 |
| *** dolphm has joined #openstack-dev | 16:42 | |
| jeblair | milner: i re-approved it, jenkins should get to work on it now | 16:45 |
| eglynn | ttx: quick question, what are the mechanics of taking a fix just merged to master and backporting it to milestone-proposed (in order to get it into E4) | 16:46 |
| eglynn | ttx: is it just a case of re-proposing the review on the milestone-proposed branch? | 16:46 |
| ttx | eglynn: http://wiki.openstack.org/GerritJenkinsGithub#Submit_Changes_in_master_to_milestone-proposed | 16:46 |
| *** dolphm has quit IRC | 16:47 | |
| milner | jeblair: great, thanks! | 16:47 |
| eglynn | ttx: thx! | 16:47 |
| *** letterj has joined #openstack-dev | 16:48 | |
| *** ChanServ sets mode: +v letterj | 16:48 | |
| *** andrewsben has quit IRC | 16:48 | |
| letterj | LinuxJedi: I'm the one that notmyname was helping with reviews.openstack.org. Can I ask a few questions? | 16:49 |
| LinuxJedi | letterj: of course | 16:49 |
| letterj | Where is launch pad do I find the identity that I want them to use? I see the identity used on review.openstack.org | 16:50 |
| LinuxJedi | letterj: if you go to your Lauchpad user page. View the source and search for "openid2.local_id" you should find it there | 16:51 |
| LinuxJedi | letterj: it is also in http://paste.openstack.org/show/5472/ for you | 16:51 |
| LinuxJedi | letterj: along with the wrong one Launchpad are sending when you login to gerrit | 16:51 |
| letterj | oh wow. thanks for the help. I was told that was just an example of how someone else did it. :-) | 16:53 |
| annegentle | letterj: LinuxJedi: I had to do the same thing, hopefully they'll see it as a bug and let us consolidate. | 16:53 |
| letterj | LinuxJedi: thanks again for your help | 16:54 |
| LinuxJedi | letterj: no problem at all, unfortunately this comes up a couple of times a week now, really wish LP would pull their finger out and fix it :( | 16:55 |
| *** jdg has joined #openstack-dev | 16:55 | |
| *** oneiroi has quit IRC | 16:55 | |
| gkotton | anyone try and run devstack as of late | 16:55 |
| *** oneiroi has joined #openstack-dev | 16:56 | |
| mjfork | gkotton: i did this AM and it failed. | 16:56 |
| *** andrewsben has joined #openstack-dev | 16:56 | |
| dtroyer | gkotton, mjfork : there is known devstack breakage with horizon's refactor, review of fix is in progress | 16:57 |
| eglynn | gkotton: I ran dtroyer's Fedora16-aware version yesterday | 16:57 |
| gkotton | mjfork: thanks | 16:57 |
| eglynn | gkotton: worked fine with a couple tweaks | 16:57 |
| gkotton | eglynn: thanks - i ran into the problems on ubuntu. i'll give the fedora a bash | 16:57 |
| *** s1rp has joined #openstack-dev | 16:57 | |
| eglynn | (e.g. creating the /opt/stack/logs dir in advance) | 16:57 |
| mjfork | gkotton: horizon related? looks like horizon was refactored and devstack not caught up | 16:58 |
| dtroyer | mjfork: correct | 16:58 |
| *** zzed has quit IRC | 16:59 | |
| *** zzed has joined #openstack-dev | 16:59 | |
| *** cjwilson has quit IRC | 17:00 | |
| *** Mkenneth has quit IRC | 17:00 | |
| *** dolphm has joined #openstack-dev | 17:00 | |
| gkotton | mjfork: yes related to horizon | 17:01 |
| *** sdague has joined #openstack-dev | 17:02 | |
| *** maploin has quit IRC | 17:02 | |
| *** oneiroi has quit IRC | 17:08 | |
| *** littleidea has quit IRC | 17:08 | |
| Kiall | mtaylor / jeblair mind me asking how you have Jenkins report build status back for the nova-tarball run? eg https://review.openstack.org/3178 | 17:14 |
| jeblair | Kiall: sure | 17:15 |
| jeblair | Kiall: for a while we ran with a few tarball jobs set to trigger on change-merged | 17:15 |
| jeblair | Kiall: we thought that was the right thing for a tarball job. since change-merged has a change associated with it, it can report back to gerrit | 17:16 |
| jeblair | Kiall: however, there's a problem with that and we stopped using change-merged: | 17:16 |
| Kiall | Right, I've got some similar jobs trigger by change-merged - would like their results posted back once they complete.. | 17:16 |
| jeblair | Kiall: the job runs with parameters that are associated with the current patchset; however since we allow gerrit to merge commits if needed, sometimes the result in the tree is the patchset as uploaded PLUS a merge commit that merges it into the tip. change-merged events lack information about the merge commit, so if you check out the git repo with only what's included, you'll get the change as the person submitted it, rather than the | 17:18 |
| *** bengrue has joined #openstack-dev | 17:18 | |
| jeblair | Kiall: that's the wrong thing to build a tarball from, so we stopped. | 17:18 |
| jeblair | Kiall: that's when we started down the ref-updated path | 17:19 |
| zaitcev | eglynn: Were you able to find an answer how to re-start Jenkins? I mean, other than "poke jpipes on IRC". | 17:19 |
| jeblair | Kiall: though you can also solve that problem in another way: have gerrit include merge commit info in the change-merged event, and have the gerrit trigger plugin pick up on that. I've actually just about finished a patch to gerrit to do the 1st part of that. | 17:19 |
| zaitcev | Or "upload the same rewiew set again" | 17:19 |
| Kiall | jeblair: Aha, interesting.. I haven't considered that.. | 17:20 |
| *** jpipes is now known as jaypipes | 17:20 | |
| Kiall | zaitcev: join the "nova" launchpad team and you seem to get access to the retrigger button.. | 17:20 |
| eglynn | zaitcev: the IRC poke to a core reviewer is only way I know how to re-trigger unfortunately | 17:20 |
| *** bencherian has joined #openstack-dev | 17:20 | |
| eglynn | Kiall: interesting ... | 17:20 |
| jeblair | Kiall: so for your actual question, er, i think we just left the defaults and it reported back. i haven't tested that behavior in my latest build -- its possible something has happened to stop that. | 17:20 |
| Kiall | (At least - I'm guessing its that! Since I somehow see the retrigger button and most people dont) | 17:20 |
| *** davidkranz has quit IRC | 17:21 | |
| Kiall | jeblair: I saw the same behaviour on an older build too, no reporting back for change-merged builds... | 17:21 |
| *** aweiss has quit IRC | 17:22 | |
| jeblair | Kiall: also, we hang out on #openstack-infra and talk about this sort of thing all day if you're interested. | 17:22 |
| Kiall | Lol - I would like to occasionally get some work done ;) | 17:22 |
| *** dubsquared has joined #openstack-dev | 17:22 | |
| *** x86brandon has joined #openstack-dev | 17:23 | |
| *** x86brandon_ has joined #openstack-dev | 17:23 | |
| *** bepernoot has joined #openstack-dev | 17:24 | |
| *** x86brandon_ has quit IRC | 17:24 | |
| *** x86brandon__ has joined #openstack-dev | 17:24 | |
| zul | mtaylor: cool | 17:24 |
| eglynn | jaypipes: can I trouble you to repeat your earlier +2 on https://review.openstack.org/#change,4602 ? | 17:25 |
| eglynn | jaypipes: the patch now has multiple approvals and is verified (finally!) by Jenkins, but no current +2 so merge is blocked | 17:25 |
| eglynn | jaypipes: also https://review.openstack.org/#change,4447 if you have time | 17:26 |
| openstackgerrit | Verification of a change to openstack/quantum failed: remove pep8 and strict lxml version from setup.py https://review.openstack.org/4235 | 17:26 |
| jaypipes | eglynn: yup, gimme 10 minutes | 17:26 |
| eglynn | jaypipes: cool, thanks! | 17:26 |
| *** s1rp has quit IRC | 17:27 | |
| *** x86brandon has quit IRC | 17:27 | |
| *** x86brandon__ is now known as x86brandon | 17:27 | |
| *** nati_ueno has joined #openstack-dev | 17:27 | |
| *** popux has quit IRC | 17:28 | |
| jaypipes | eglynn: off to the pit of despair it goes... | 17:28 |
| eglynn | jaypipes: thanks, I'll re-propose patch on milestone-proposed once it's merged to master | 17:29 |
| eglynn | (so as to get it into E4) | 17:29 |
| jaypipes | eglynn: yup, same for 4447 | 17:30 |
| eglynn | jaypipes: cool | 17:30 |
| jdg | Anybody able to approve https://review.openstack.org/#change,4403 for me? It's been sitting waiting for a while | 17:30 |
| *** derekh has quit IRC | 17:31 | |
| *** hhoover has left #openstack-dev | 17:31 | |
| jdg | jaypipes: thanks! | 17:32 |
| *** shevek__ has quit IRC | 17:32 | |
| jaypipes | jdg: least I could do :) | 17:32 |
| jaypipes | jdg: not nova-core, though... so only a slight comfort :) | 17:32 |
| jdg | jaypipes: Progress is good no matter what :) | 17:33 |
| *** adjohn has joined #openstack-dev | 17:33 | |
| *** hashar has quit IRC | 17:34 | |
| openstackgerrit | Verification of a change to openstack/quantum failed: remove pep8 and strict lxml version from setup.py https://review.openstack.org/4235 | 17:35 |
| *** dalang has joined #openstack-dev | 17:35 | |
| zykes- | Kiall: around ? | 17:36 |
| Kiall | yea | 17:36 |
| zykes- | what you think of the idea to | 17:36 |
| zykes- | couchbase > python > powerdns as prim dns ? | 17:37 |
| Kiall | The same (constructive) criticism as last time.. Why introduce the complexity of another service like Couchbase? | 17:38 |
| zykes- | why else is the use of using powerdns if you need to put bind in front of it ? | 17:38 |
| jaypipes | dtroyer: quick heads up.. I removed keystone from being affected by bug 942684 and updated the description to reflect the work you are doing in devstack to address the devstack pieces of http://etherpad.openstack.org/keystone-admin-config | 17:38 |
| uvirtbot` | Launchpad bug 942684 in devstack "devstack should only use service_user/service_password, not admin_token" [Critical,In progress] https://launchpad.net/bugs/942684 | 17:38 |
| Kiall | PowerDNS can talk the the existing databases people deploy with nova, reusing those makes sense - unless there is a really good reason for deploying a new service... | 17:39 |
| openstackgerrit | Verification of a change to openstack/quantum failed: Split out pip requires and aligned tox file. https://review.openstack.org/4699 | 17:39 |
| zykes- | yeah, but it doesn't make sense to use powerdns when you need to have a bind one in front of it | 17:39 |
| zykes- | feels utterly silly | 17:39 |
| dtroyer | jaypipes: thanks. It's actually in 942983 now. | 17:40 |
| Kiall | The reason I suggested BIND in front on PowerDNS is that a DB backed public DNS server will be resource heavy... | 17:40 |
| Kiall | You end up with the advantages of PowerDNS, but none of its drawbacks | 17:40 |
| zykes- | hmm | 17:40 |
| jaypipes | dtroyer: ah.. ok. want me to mark the other a dup? | 17:41 |
| dtroyer | no, I think they're different. | 17:41 |
| dtroyer | different details? | 17:41 |
| *** vincentricci has joined #openstack-dev | 17:42 | |
| jaypipes | dtroyer: well, I think the stuff that related to Keystone (me mistaking the changes in the handling of service tokens) turned out to be Invalid, so the only thing left is the same as the other bug :) | 17:42 |
| dtroyer | ah, ok, dupe it then | 17:43 |
| jaypipes | dtroyer: done :) | 17:43 |
| *** jakedahn has quit IRC | 17:43 | |
| *** vincentricci has left #openstack-dev | 17:43 | |
| *** joesavak has quit IRC | 17:43 | |
| *** jakedahn has joined #openstack-dev | 17:44 | |
| *** jakedahn has quit IRC | 17:45 | |
| *** Mandell has joined #openstack-dev | 17:46 | |
| *** zigo has joined #openstack-dev | 17:47 | |
| *** dolphm has quit IRC | 17:54 | |
| *** jdurgin has joined #openstack-dev | 17:55 | |
| dtroyer | For those having trouble with horizon and devstack this morning, the fix has been merged into devstack | 17:56 |
| *** troytoman-away has quit IRC | 17:57 | |
| *** troytoman-away has joined #openstack-dev | 17:57 | |
| *** troytoman-away has quit IRC | 17:58 | |
| *** troytoman-away has joined #openstack-dev | 17:58 | |
| *** dolphm has joined #openstack-dev | 17:59 | |
| *** ches has quit IRC | 18:00 | |
| *** ches has joined #openstack-dev | 18:00 | |
| *** hashar has joined #openstack-dev | 18:02 | |
| eglynn | jaypipes: those 2 patches re-submitted on milestone-proposed ... https://review.openstack.org/4702 https://review.openstack.org/4704 | 18:08 |
| jaypipes | eglynn: cheers | 18:09 |
| *** dolphm has quit IRC | 18:09 | |
| *** troytoman-away has quit IRC | 18:10 | |
| *** troytoman-away has joined #openstack-dev | 18:10 | |
| *** bhall has joined #openstack-dev | 18:11 | |
| *** bhall has quit IRC | 18:11 | |
| *** bhall has joined #openstack-dev | 18:11 | |
| *** jakedahn has joined #openstack-dev | 18:12 | |
| *** zaitcev has quit IRC | 18:13 | |
| jdg | Can anyone confirm that summit registration for folks without any invite code opens up tomorrow? | 18:16 |
| *** dolphm has joined #openstack-dev | 18:18 | |
| jaypipes | jdg: see my PM... | 18:20 |
| annegentle | jdg: confirmed, and if you have an invite, use it today | 18:20 |
| *** darraghb has quit IRC | 18:21 | |
| jdg | annegentle: I've been asked by a customer how they can attend, already used my invite last week :) | 18:22 |
| *** kbringard has quit IRC | 18:22 | |
| *** kbringard has joined #openstack-dev | 18:23 | |
| *** novas0x2a|laptop has joined #openstack-dev | 18:24 | |
| *** troytoman-away is now known as troytoman | 18:27 | |
| *** reed has joined #openstack-dev | 18:27 | |
| *** zigo has quit IRC | 18:28 | |
| termie | heckj: did we do the thing with the cutting a release? | 18:29 |
| *** adjohn has quit IRC | 18:34 | |
| *** dolphm has quit IRC | 18:34 | |
| *** dolphm has joined #openstack-dev | 18:35 | |
| *** Vek has joined #openstack-dev | 18:37 | |
| *** Vek has quit IRC | 18:37 | |
| *** Vek has joined #openstack-dev | 18:37 | |
| eglynn | vishy: can I trouble you with https://review.openstack.org/#change,4706 ? | 18:41 |
| eglynn | vishy: just the same nova rate limiting fix you +2'd yesterday, resubmitted on milestone-proposed in order to get it into E4 | 18:42 |
| jdg | vishy: While folks are making requests maybe https://review.openstack.org/#change,4403 as well (one line change to add lun number to volume update) | 18:46 |
| *** novas0x2a|laptop has quit IRC | 18:47 | |
| *** novas0x2a|laptop has joined #openstack-dev | 18:47 | |
| *** dolphm has quit IRC | 18:48 | |
| *** dolphm has joined #openstack-dev | 18:49 | |
| *** pvo has quit IRC | 18:49 | |
| *** pvo has joined #openstack-dev | 18:49 | |
| *** ChanServ sets mode: +v pvo | 18:49 | |
| *** andrewbogott is now known as andrewbogott_afk | 18:52 | |
| *** Ryan_Lane has joined #openstack-dev | 18:52 | |
| *** bepernoot has quit IRC | 18:52 | |
| vishy | eglynn: it didn't make it in? | 18:53 |
| *** dtroyer has quit IRC | 18:53 | |
| *** dtroyer has joined #openstack-dev | 18:54 | |
| openstackgerrit | Verification of a change to openstack/glance failed: Allow region selection when using V2 keystone https://review.openstack.org/4350 | 18:54 |
| *** mjfork has quit IRC | 18:57 | |
| *** mjfork has joined #openstack-dev | 18:57 | |
| eglynn | vishy: nope, only got into master after the milestone-proposed branch was cut | 18:57 |
| *** dtroyer_ has joined #openstack-dev | 19:01 | |
| *** cp16net has quit IRC | 19:01 | |
| jakedahn | jaypipes: https://review.openstack.org/#change,3791 | 19:02 |
| jakedahn | good call | 19:03 |
| *** jakedahn has quit IRC | 19:03 | |
| *** cp16net has joined #openstack-dev | 19:04 | |
| *** dtroyer has quit IRC | 19:04 | |
| *** dtroyer_ is now known as dtroyer | 19:04 | |
| *** torgomatic has joined #openstack-dev | 19:04 | |
| *** littleidea has joined #openstack-dev | 19:05 | |
| jaypipes | jeblair: ping | 19:06 |
| jaypipes | eglynn: ok, I'm officially stumped why that regional endpoint patch won't pass the integration-devstack-vm gate :( Keeps failing with this: https://jenkins.openstack.org/job/gate-integration-tests-devstack-vm/1952/console | 19:08 |
| jaypipes | eglynn: and I'd really like to get that patch into milestone-proposed... | 19:09 |
| jaypipes | eglynn: any ideas? | 19:09 |
| *** eglynn has quit IRC | 19:10 | |
| *** andrewbogott_afk is now known as andrewbogott | 19:12 | |
| *** apevec has quit IRC | 19:17 | |
| jaypipes | jeblair: gonna run and grab some coffee, but my request is this: it seems that because I am a member of OpenStack admins (which is the owner of the keystone-core and horizon-core teams on Launchpad) that I have +2 ability on both keystone and horizon. I shouldn't have that ability, nor do I particularly want that ability, and I'm sure it would make termie's day if you could remove me from +2-ability in those projects. | 19:17 |
| jaypipes | Thanks. | 19:17 |
| *** zaitcev has joined #openstack-dev | 19:17 | |
| *** x86brandon has quit IRC | 19:19 | |
| *** aweiss has joined #openstack-dev | 19:20 | |
| *** vincentricci has joined #openstack-dev | 19:27 | |
| *** mnewby has joined #openstack-dev | 19:29 | |
| *** eglynn has joined #openstack-dev | 19:31 | |
| *** nati_ueno has quit IRC | 19:31 | |
| bengrue | anyone mind if I grab https://bugs.launchpad.net/keystone/+bug/942986 | 19:31 |
| uvirtbot` | Launchpad bug 942986 in keystone "auth_token needs logging enabled" [High,Confirmed] | 19:31 |
| jaypipes | bengrue: check with anotherjesse ... he may be working in there.. | 19:32 |
| *** mattray has joined #openstack-dev | 19:32 | |
| eglynn | jaypipes: sorry missed your ping earlier | 19:33 |
| eglynn | jaypipes: so that devstack failure ... | 19:33 |
| bcwaldon | bengrue: anotherjesse is working here -> https://review.openstack.org/#change,4675 | 19:33 |
| bcwaldon | bengrue: confict-central | 19:33 |
| eglynn | jaypipes: nova image-list failing with 500 ... are the nova and glance logs preserved after the test run completes? | 19:34 |
| bengrue | oh, oh I see. | 19:34 |
| heckj | bengrue: I don't know that jesse is getting logging in there though | 19:34 |
| bengrue | mmm. | 19:34 |
| bengrue | is there any project that's not being gated on human interaction right now that needs some lovin'? | 19:35 |
| bcwaldon | heckj: he isn't addressing that bug, I was just pointing out a sure confclit | 19:35 |
| bcwaldon | whoops | 19:35 |
| heckj | bcwaldon - yeah, definitely | 19:35 |
| heckj | bengrue: I'd love to see the logging in there, but be aware of the impending conflict ... | 19:36 |
| jeblair | jaypipes: it's likely that you slipped in there in an earlier version of the sync script. i'll remove you from those groups now. the next version of the script (to be deployed soon) should correctly remove people from groups. | 19:36 |
| jaypipes | eglynn: hmm... | 19:36 |
| *** jakedahn has joined #openstack-dev | 19:37 | |
| jeblair | jaypipes, eglynn: https://jenkins.openstack.org/job/gate-integration-tests-devstack-vm/1952/artifact/logs/syslog.txt | 19:38 |
| jeblair | the syslog is preserved for each devstack gate job | 19:39 |
| jeblair | (look under "Build Artifacts" on the page for the build) | 19:39 |
| eglynn | jeblair: nice! thx | 19:39 |
| * eglynn looking ... | 19:40 | |
| jaypipes | eglynn, jeblair: related? https://review.openstack.org/#change,4698 | 19:40 |
| *** sbroeker has joined #openstack-dev | 19:41 | |
| *** Lxu has quit IRC | 19:41 | |
| sbroeker | I am having trouble with "git review" | 19:42 |
| sbroeker | I get the following error message: | 19:42 |
| sbroeker | cp: :hooks/commit-msg: No such file or directory" | 19:43 |
| sbroeker | Any ideas? | 19:43 |
| eglynn | jaypipes: wasn't Stuart's patch failing devstack before 4698 went into nova? | 19:43 |
| jaypipes | eglynn: yes... I suppose we can rekick that job and see if it makes a diff... | 19:44 |
| jeblair | sbroeker: can you run "git-review -s -v" and copy the output to paste.openstack.org ? | 19:44 |
| jaypipes | eglynn: I retriggered it... | 19:44 |
| eglynn | jaypipes: cool | 19:44 |
| *** anotherjesse1 has joined #openstack-dev | 19:44 | |
| jaypipes | eglynn: if it works, fine, if not, I think it may have to do with other devstack/keystone things currently being worked on. | 19:44 |
| *** bencherian_ has joined #openstack-dev | 19:45 | |
| *** bencherian has quit IRC | 19:45 | |
| *** bencherian_ is now known as bencherian | 19:45 | |
| *** adjohn has joined #openstack-dev | 19:45 | |
| *** pixelbeat has quit IRC | 19:46 | |
| eglynn | jaypipes: I'll spin up a fresh devstack here with Stuart's patch applied to glance, do a bit a digging to see if I can twigg what's going on ... | 19:46 |
| jaypipes | eglynn: k, thx | 19:47 |
| sbroeker | jeblair: I did the paste #5509. | 19:48 |
| openstackgerrit | Verification of a change to openstack/glance failed: Allow region selection when using V2 keystone https://review.openstack.org/4350 | 19:50 |
| anotherjesse1 | dtroyer / jaypipes - the service_pass/protocol/host/port/url - I think those can be removed | 19:50 |
| anotherjesse1 | from auth_token.py - they were optional, so that you can deploy auth_token.py as a stand-alone service (eg, it turned it into an auth-ing proxy) | 19:51 |
| jeblair | sbroeker: what does "git-review --version" say? | 19:51 |
| *** paulormg has quit IRC | 19:51 | |
| sbroeker | jeblair: git-review version 1.14 | 19:52 |
| jeblair | mtaylor: it looks like your fix was applied after 1.14; time to cut 1.15? | 19:54 |
| eglynn | jaypipes: just tried on a day-old devstack, nova image-list failing in the keystone auth_token middleware: http://fpaste.org/oUuK/ | 19:56 |
| eglynn | jaypipes: actually scratch that ... I think dprince has already fixed the KeyError on the json.loads(data)["access"]["token"]["id"] nested lookup | 19:57 |
| bcwaldon | eglynn: should be fixed as of now | 19:57 |
| bcwaldon | eglynn: also, can't see that paste | 19:57 |
| eglynn | bcwaldon: here it is http://www.pastie.org/3490754 but yeah, you're right already fixed | 19:59 |
| * eglynn spinning up a fresh devstack ... | 19:59 | |
| jeblair | sbroeker: http://pypi.python.org/pypi/git-review/1.15 | 20:01 |
| jeblair | sbroeker: try installing that version; it has a fix for what I believe is the problem you are encountering. let me knwo if it doesn't work | 20:01 |
| *** anotherjesse2 has joined #openstack-dev | 20:03 | |
| anotherjesse2 | jaypipes: around? | 20:06 |
| *** anotherjesse1 has quit IRC | 20:07 | |
| *** nati_ueno has joined #openstack-dev | 20:10 | |
| jaypipes | anotherjesse: yep, back from coffee | 20:11 |
| *** x86brandon has joined #openstack-dev | 20:11 | |
| *** cp16net has quit IRC | 20:11 | |
| anotherjesse2 | jaypipes: the "service" stuff was for a mode that wasn't used - http://pastie.org/3490806 | 20:11 |
| *** cp16net has joined #openstack-dev | 20:12 | |
| anotherjesse2 | running in http auth proxy mode - if you want to do that it should be the auth_token.py middleware and then an additional auth_proxy.py (don't put both features in a single wsgi app) | 20:12 |
| jaypipes | anotherjesse: yes, no disagreement there :) I was just saying that I think s/admin/service/g, but it's not a huge deal. | 20:13 |
| anotherjesse2 | jaypipes: if this review lands there won't be overlapping meaning of "service_*" | 20:13 |
| anotherjesse2 | so - we could potentially go with service | 20:13 |
| jaypipes | anotherjesse: yeah, that's all I was suggesting, but it's not a biggie | 20:14 |
| jaypipes | anotherjesse: a more important comment from me was about removing the token entirely from the conf... | 20:15 |
| *** torgomatic has quit IRC | 20:17 | |
| anotherjesse2 | jaypipes: I think we should mark it as deprecated but removing it might break people | 20:17 |
| *** ameade_ is now known as ameade | 20:17 | |
| anotherjesse2 | jaypipes / dtroyer: also replied to some of your comments in https://review.openstack.org/#change,4675 - still looking at the bottom two comments | 20:18 |
| jaypipes | anotherjesse: actually, keeping it in there is breaking people :) Since people have been putting in what they think is a service token (long-lived token), which is now not possible to generate... | 20:18 |
| anotherjesse2 | the thing breaking is that auth_token.py (still) will mark _admin_token to None if the user's token is invalid... | 20:18 |
| anotherjesse2 | I agree with improving the situation - still working through all the details - I might agree with you in the end ;) | 20:19 |
| jaypipes | anotherjesse: yup, agree, it's a tricky one.. | 20:19 |
| *** jog0 has joined #openstack-dev | 20:20 | |
| *** aweiss has quit IRC | 20:20 | |
| jaypipes | anotherjesse: the comment of mine that begins with Two things: is the important one IMHO | 20:21 |
| anotherjesse2 | replying to that one now | 20:21 |
| jaypipes | anotherjesse: it's the source of the bug and the reason the 503 raise doesn't work properly. | 20:21 |
| jaypipes | kk | 20:21 |
| *** creiht has joined #openstack-dev | 20:23 | |
| *** ChanServ sets mode: +v creiht | 20:23 | |
| *** aweiss has joined #openstack-dev | 20:23 | |
| anotherjesse2 | jaypipes: replied | 20:24 |
| anotherjesse2 | jaypipes: I commented on where I think the issue is here: https://review.openstack.org/#patch,sidebyside,4675,12,keystone/middleware/auth_token.py | 20:26 |
| jaypipes | anotherjesse: ok, so yeah... the property is accessed in validate_user_token(), true, but the logic that checks for not None (after the property will return None) is lacking from validate_user_token(), which means a X-Auth-Token: None is sent through...and not a 503 returned | 20:27 |
| anotherjesse2 | yep! | 20:27 |
| anotherjesse2 | agreed - hence I left the comment in admin_token method saying there is still a bug in here | 20:27 |
| jaypipes | anotherjesse: so I think it might be better to make the property into a pure function and have it raise a 503 and do the retry itself? | 20:27 |
| anotherjesse2 | jaypipes: probably | 20:28 |
| jaypipes | anotherjesse: and only do the retry once in the main __call__ method? | 20:28 |
| anotherjesse2 | jaypipes: btw, an issue that hit me for a bit. if you do glance -A foo index; it will work if you have environmental variables set correctly (user/pass/tenant) - because it will hit glance with invalid token, which glance gets told is invalid, then glance client will talk to keystone and get a new token with the env variables and then hit glance again and succeed | 20:29 |
| jaypipes | anotherjesse: i.e. check for admin_token in __call__. If not existing, call self._get_admin_token(), with _get_admin_token() raising a 503 itself after a single retry to deal with expired admin token? | 20:29 |
| *** adjohn has quit IRC | 20:29 | |
| anotherjesse2 | jaypipes: I'm thinking something like that! | 20:29 |
| jaypipes | anotherjesse: yeah, there's some overhaul needed in glance CLI client... | 20:30 |
| jaypipes | anotherjesse: and better ways of logging debug messages in the glance client | 20:30 |
| dprince | anotherjesse2/jaypipes: I'm a fan of switching from admin --> service for the config names. | 20:30 |
| anotherjesse2 | I thought it was letting things through with the A**BAG token | 20:30 |
| jaypipes | anotherjesse: :) | 20:30 |
| anotherjesse2 | turns out the client was smart | 20:30 |
| jaypipes | anotherjesse: no, have to use the super-special-secret "F**ktard" password. | 20:30 |
| * jaypipes goes off to write easter egg into python-glanceclient... | 20:31 | |
| jaypipes | eglynn: retrigger of the regional endpoint job failed again.. | 20:32 |
| eglynn | jaypipes: bummer! | 20:32 |
| eglynn | jaypipes: I'll continue digging ... | 20:33 |
| jaypipes | eglynn: yeah I know... but I have a sneaking suspicion the root cause has something to do with what anotherjesse and others are in the process of working on | 20:33 |
| eglynn | jaypipes: cool | 20:33 |
| jaypipes | anotherjesse: OK, so we are definitely on the same wavelength re: the middleware. Good to know. | 20:34 |
| anotherjesse2 | jaypipes: will continue on it after lunch | 20:34 |
| anotherjesse2 | I'm waiting on dolphm to respond to my comment on patch 12 | 20:34 |
| jaypipes | anotherjesse: rock on. | 20:35 |
| jaypipes | anotherjesse: good to hear they're letting you have lunch now. | 20:35 |
| jaypipes | :P | 20:35 |
| *** dneary has joined #openstack-dev | 20:35 | |
| ayoung | heckj, is the current master of keystone client compatable with the current master of keystone? | 20:36 |
| *** andrewsben has quit IRC | 20:38 | |
| heckj | ayoung: yes, we test keystone agressively against keystone-client with every potential update. Note: we don't do the reverse, so we can introduce breaks with changes to keystone-client (a hole we need to fill) | 20:39 |
| ayoung | heckj, so I finally ran the keystoneclient code against ldap Identity and found I missed user_by_name | 20:39 |
| ayoung | I'll be posting a patch for that shortly | 20:40 |
| heckj | ayoung: excellent! | 20:40 |
| ayoung | will I need a ticket for that? | 20:40 |
| ayoung | this time I promise to run pep8 on it prior to submitting, too | 20:40 |
| heckj | ayoung: a bug never hurts, but isn't required. | 20:42 |
| ayoung | heckj, OK. I'll ticket it. | 20:43 |
| *** davidkranz has joined #openstack-dev | 20:46 | |
| *** Kiall has quit IRC | 20:47 | |
| *** bepernoot has joined #openstack-dev | 20:50 | |
| *** Kiall has joined #openstack-dev | 20:50 | |
| *** andrewsben has joined #openstack-dev | 20:53 | |
| *** pixelbeat has joined #openstack-dev | 20:55 | |
| *** adjohn has joined #openstack-dev | 20:56 | |
| *** bepernoot has quit IRC | 20:58 | |
| YorikSar | bcwaldon: around? | 20:59 |
| *** nati_ueno has quit IRC | 21:03 | |
| *** cmagina has quit IRC | 21:07 | |
| *** cmagina has joined #openstack-dev | 21:07 | |
| *** dprince has quit IRC | 21:11 | |
| Daviey | vishy / devcamcar: Do one of you want to approve https://review.openstack.org/4690 ? you've both given it a +1 | 21:13 |
| devcamcar | Daviey: looking | 21:13 |
| devcamcar | ahh, yes, i meant to | 21:14 |
| devcamcar | flipped the bit | 21:14 |
| Daviey | ta! | 21:14 |
| *** dneary has quit IRC | 21:20 | |
| *** joesavak has joined #openstack-dev | 21:21 | |
| *** joesavak has quit IRC | 21:21 | |
| *** rkukura has quit IRC | 21:22 | |
| *** stuntmachine has quit IRC | 21:25 | |
| *** medberry is now known as med_ | 21:27 | |
| openstackgerrit | Verification of a change to openstack/swift failed: Add support for URLs and absolute paths in staticweb CSS listings. https://review.openstack.org/4442 | 21:30 |
| bcwaldon | YorikSar: yep, what's up | 21:30 |
| notmyname | mtaylor: how do I unstuck that ^ (it's not a swift error, it's an error with the git server) | 21:31 |
| notmyname | mtaylor: ah. I found a "retrigger" button in jenkins | 21:34 |
| *** milner has quit IRC | 21:37 | |
| *** PotHix has quit IRC | 21:40 | |
| *** nati_ueno has joined #openstack-dev | 21:41 | |
| ttx | jaypipes: don't forget to readjust the Glance E4 bugs list before EOD so that I know what we are blocking on | 21:43 |
| jaypipes | ttx: yup, will do right now. thx for the reminder | 21:44 |
| jaypipes | bcwaldon, anotherjesse, mtaylor: started new glance CLI and client package: https://github.com/jaypipes/python-glanceclient | 21:45 |
| jaypipes | obviously long way to go, but it's a start... looking to Folsom to get alignment with other cores. | 21:45 |
| jeblair | jaypipes: would you like that to be in jenkins/gerrit? | 21:46 |
| ttx | jeblair: not for essex | 21:47 |
| jaypipes | jeblair: yes, but like ttx says, for folsom.. | 21:47 |
| ttx | I'd rather not change the set of deliverables at this point | 21:47 |
| openstackgerrit | Verification of a change to openstack/nova failed: Raise 409 when rescuing instance in RESCUE mode https://review.openstack.org/4599 | 21:47 |
| jaypipes | ttx: I wasn't suggesting we would change deliverables | 21:47 |
| jeblair | we don't have to wait until it's released to start using the project infrastructure; in general it's better to have projects move into our infrastructure earlier rather than later i think | 21:48 |
| bcwaldon | jaypipes: awesome, thanks for starting on that | 21:48 |
| ttx | vishy: suggest we add bug 943293 to E4, will do the MP backport tomorrow | 21:49 |
| uvirtbot` | Launchpad bug 943293 in nova "Missing rootwrap filters for newly-introduced run_as_root commands" [High,Fix committed] https://launchpad.net/bugs/943293 | 21:49 |
| jaypipes | jeblair: that's fine... just saying it's not a priority and it won't be replacing the python-glance package in PyPI for the Essex release | 21:50 |
| jeblair | jaypipes: cool, can you file an openstack-ci bug about adding it to jenkins/gerrit when you're ready? | 21:52 |
| *** nati_ueno has quit IRC | 21:53 | |
| Daviey | jaypipes: you almost gave me a heart attack then.. i thought you were going to try to target that new thing at essex. | 21:53 |
| Daviey | :) | 21:53 |
| jaypipes | jeblair: will do | 21:53 |
| jaypipes | Daviey: lol, not a chance | 21:54 |
| *** zns has joined #openstack-dev | 21:54 | |
| anotherjesse2 | jaypipes: http://pastie.org/3491595 <- trying to get logging working | 21:54 |
| *** zns has left #openstack-dev | 21:55 | |
| zaitcev | I can tell you that's not going to work | 21:55 |
| zaitcev | Meh, I had a working logging patch somewhere.... in one of my repos | 21:55 |
| *** joesavak has joined #openstack-dev | 21:56 | |
| anotherjesse2 | zaitcev: yeah, I have no clue what is going on ;( | 21:56 |
| zaitcev | anotherjesse2: Anyway, basically, look how Swift does it. It has a helper loke get_logger() which actually works. You need some kind of... like set some flags or whatever. Just steal from there | 21:56 |
| anotherjesse2 | k | 21:56 |
| *** aweiss has quit IRC | 21:56 | |
| eglynn | jaypipes: dunno if this is the *same* issue that causing Jenkins devstack integ test to fail for Stuart's region patch | 21:56 |
| eglynn | jaypipes: but I'm seeing nova image-list fail in the strict creds completeness test he added | 21:57 |
| eglynn | jaypipes: but I'm seeing nova image-list fail in the strict creds completeness test he added jaypipes: failure http://www.pastie.org/3491587 | 21:57 |
| eglynn | jaypipes: corresponding code change https://review.openstack.org/#patch,sidebyside,4350,4,glance/common/auth.py lines 81-82 | 21:57 |
| jaypipes | anotherjesse: before using the swift stuff... what exactly are you experiencing? that middleware should be loaded into the pipeline *after* logging is configured. | 21:57 |
| zaitcev | Wait, I found it | 21:57 |
| anotherjesse2 | jaypipes: trying to make it so I can log critical exceptions in the auth_token middleware | 21:58 |
| zaitcev | anotherjesse2: http://people.redhat.com/zaitcev/tmp/x.diff | 21:58 |
| zaitcev | brb | 21:58 |
| jaypipes | zaitcev: all that code should already have been run in setup_logging() before the pipeline even initalizes the middleware... | 21:59 |
| jaypipes | anotherjesse: can you verify that setup_logging() has been called? | 21:59 |
| jaypipes | eglynn: looking... | 22:00 |
| bcwaldon | jaypipes: if so, then what jesse has in that paste is fine, right? | 22:00 |
| bcwaldon | jaypipes: RE logging | 22:00 |
| jaypipes | bcwaldon: should be... depends on the code that comes before the middleware initialization... setup_logging() needs to be called before load_paste_app(). | 22:01 |
| *** andrewsmedina has quit IRC | 22:01 | |
| jaypipes | eglynn: but didn't that patch from stu with the strict creds already hit trunk? :( | 22:01 |
| *** andrewsben has quit IRC | 22:01 | |
| bcwaldon | jaypipes: but the middleware is getting a different logger than the setup would do | 22:02 |
| jaypipes | eglynn: or did he roll these patches into one? .. | 22:02 |
| bcwaldon | jaypipes: since the setup would be for glance or nova, but he's using a keystone logger | 22:02 |
| *** troytoman is now known as troytoman-away | 22:02 | |
| eglynn | jaypipes: yep, I see, he just moved that line and added strategy to the required creds | 22:02 |
| jaypipes | bcwaldon: in setup_logging(), try putting logger.propogate = True | 22:03 |
| *** joesavak has quit IRC | 22:03 | |
| jaypipes | eglynn: so we need to supply strategy now? .. | 22:04 |
| jaypipes | eglynn: hmm... not sure if that is very user-friendly :) | 22:04 |
| jaypipes | eglynn: if auth_url is specified probably would be best to default to keystone strategy. thoughts? | 22:04 |
| eglynn | jaypipes: yep, but its actually failing on a missing passwd | 22:04 |
| eglynn | MissingCredentialError: Missing required credential: password | 22:04 |
| eglynn | jaypipes: despite OS_PASSWORD being set in my env | 22:05 |
| * eglynn will dig some more ... | 22:05 | |
| jaypipes | k, appreciated. | 22:05 |
| jaypipes | I'll ping stuart and see what the status is on https://bugs.launchpad.net/glance/+bug/920386. I could have sworn we'd gotten that into trunk already... | 22:06 |
| uvirtbot` | Launchpad bug 920386 in glance "Client side SSL related variables" [Medium,Triaged] | 22:06 |
| *** flaviamissi has quit IRC | 22:06 | |
| *** troytoman-away is now known as troytoman | 22:09 | |
| *** vricci has joined #openstack-dev | 22:10 | |
| *** vincentricci has quit IRC | 22:11 | |
| *** vricci is now known as vincentricci | 22:11 | |
| mtaylor | jaypipes: I'm making patches for you alreday... | 22:11 |
| jaypipes | mtaylor: ah, ok :) | 22:12 |
| mtaylor | jaypipes: no reason to not get you on all the latest hotness right off the bad | 22:13 |
| mtaylor | bat | 22:13 |
| jaypipes | mtaylor: k | 22:14 |
| *** aweiss has joined #openstack-dev | 22:14 | |
| *** sbroeker has quit IRC | 22:14 | |
| mtaylor | jaypipes: https://github.com/jaypipes/python-glanceclient/pull/1 | 22:15 |
| jaypipes | mtaylor: merged | 22:17 |
| jaypipes | jeblair: ^^ in case you'd already started on that... | 22:17 |
| openstackgerrit | Verification of a change to openstack/nova failed: Be consistent with disabling periodic tasks. https://review.openstack.org/4671 | 22:17 |
| jeblair | jaypipes: i hadn't; thanks. | 22:18 |
| *** sandywalsh has quit IRC | 22:18 | |
| *** ayoung is now known as ayoung-afk | 22:18 | |
| mtaylor | jaypipes: just running "tox" in your source dir will do all of the virtualenv testing you've ever wanted (tox -v will be verbose about it) | 22:18 |
| jaypipes | mtaylor: rock on. that will be especially useful when tests are added. :P | 22:19 |
| mtaylor | yes. and when they don't import keystoneclient :) | 22:19 |
| jaypipes | mtaylor: yeah, yeah :) | 22:19 |
| eglynn | jaypipes: one difference is that Stuart is calling KeystoneStrategy.check_auth_params() eagerly from __init__ | 22:19 |
| eglynn | jaypipes: whereas previously this was called from the authenticate() method | 22:19 |
| eglynn | jaypipes: may be as simple as premature enforcement of required creds | 22:19 |
| eglynn | jaypipes: the corresponding nova logic hasn't changed since 2011-09-10 | 22:19 |
| jaypipes | eglynn: hmmm... looks like you are on the right trail. | 22:20 |
| jaypipes | eglynn: care to push a change to his patch and see if that is the right issue? | 22:20 |
| notmyname | what is maru newby's nick in irc? | 22:20 |
| bhall | mnewby? | 22:21 |
| eglynn | jaypipes: yep, I'll spin up another clean devstack to compare | 22:21 |
| notmyname | mnewby: ping | 22:22 |
| notmyname | bhall: thanks | 22:22 |
| bhall | notmyname: np | 22:23 |
| notmyname | mtaylor: ping (about testing with tox) | 22:24 |
| *** lloydde has quit IRC | 22:27 | |
| *** cjwilson has joined #openstack-dev | 22:27 | |
| mtaylor | notmyname: pong | 22:28 |
| notmyname | mtaylor: I'm loking at the tx patch proposed to swift | 22:29 |
| *** torgomatic has joined #openstack-dev | 22:29 | |
| notmyname | mtaylor: anything other than `[sudo] tox` to use it (and test this patch)? | 22:29 |
| mtaylor | nope. that should do everything you've ever wanted | 22:29 |
| mtaylor | notmyname: if you want to run just a single environment, you can do "tox -epy27" or "tox -epy26" ... by default it does both | 22:30 |
| notmyname | mtaylor: ya, I just ran it. that's awesome | 22:32 |
| *** andrewbogott is now known as andrewbogott_afk | 22:32 | |
| mtaylor | sweet! glad you like it. (I was so happy when I jkoelker showed it to me) | 22:32 |
| *** sandywalsh has joined #openstack-dev | 22:32 | |
| notmyname | mtaylor: so my understanding is that you can set up the .ini file for multiple environments, right? like py26 in lucid, py27 in natty, and py26 with packages from precise, etc? | 22:33 |
| notmyname | "in lucid" etc means "the package versions from that distro" | 22:33 |
| notmyname | so eg, py26 with different version of webob and eventlet | 22:33 |
| mtaylor | notmyname: well, it doesn't really undersand natty/precise or whatnot - but you could totally do that if you set up a requirements file with info about those distros | 22:34 |
| mtaylor | notmyname: totally. I hadn't thought of doing that | 22:34 |
| notmyname | ya, that's what I meant | 22:34 |
| mtaylor | but I like the idea | 22:34 |
| notmyname | so for example, get the dependencies list and then test with each version of python for each version of the packages that ship with different distros | 22:34 |
| *** mfer has joined #openstack-dev | 22:35 | |
| notmyname | eg rhel versions of dependencies, lucid version of dependencies, and even "company specific" versions of dependencies (eg we have packaged later versions of some things like webob and eventlet) | 22:35 |
| mtaylor | it would take a little bit of work to set that up - tox installs from pip ... but we could completely make a file like "pip-requires-lucid" and put the package versions in there by hand | 22:35 |
| notmyname | that's really cool | 22:36 |
| *** hashar has quit IRC | 22:36 | |
| jeblair | notmyname, mtaylor: keep in mind that even packages as shipped by a distro may not match (or even be available on) pypi because of disto-supplied patches (backported fixes, etc) | 22:37 |
| jeblair | (but perhaps you could get close enough for your goals) | 22:38 |
| notmyname | jeblair: ya. I'm not going to go out and build all these configs (I'm really only interested in py26 on lucid right now), but the idea of being able to do that is really good | 22:40 |
| *** rkukura has joined #openstack-dev | 22:40 | |
| notmyname | and useful to the project overall | 22:40 |
| mtaylor | ++ | 22:40 |
| notmyname | mtaylor: any idea why the proposed patch uses "nosetests" in the tox.ini instead of our .unittests? | 22:41 |
| notmyname | https://review.openstack.org/#change,3602 | 22:41 |
| *** markvoelker has quit IRC | 22:41 | |
| mtaylor | notmyname: he was probably was just copying what I was doing in other projects | 22:41 |
| notmyname | mtaylor: they have test runners, too, right? | 22:42 |
| blamar | bcwaldon: 'glance add', should you be able to specify an id? it seems to be always randomly generate one for me even when I give it one | 22:42 |
| mtaylor | notmyname: many of them don't need them, so can use nose directly | 22:42 |
| mtaylor | nova is the only one remaining that can't use nova by itself | 22:42 |
| notmyname | ok. swift is the same | 22:42 |
| mtaylor | can't use nose by itself | 22:42 |
| notmyname | same == can use nose by itself | 22:42 |
| mtaylor | and we've almost got that one fixed | 22:42 |
| mtaylor | awesome | 22:42 |
| notmyname | ok, so you'd rather see it use nose directly rather than the test runner? | 22:43 |
| mtaylor | it doesn't bother me to have it call .unittests instead if you'd prefer - since your .unittests really just calls nose without modification | 22:43 |
| notmyname | also, the tox patch doesn't call the .functests runner (or the equivalent nose commands) | 22:43 |
| mtaylor | it's getting rid of boiler-plate/copied custom test runner code that I really want to see happen | 22:43 |
| *** aweiss has quit IRC | 22:43 | |
| notmyname | ya, if tox can replace our .*tests that's cool | 22:44 |
| mtaylor | do you think that both unittests and functest should be called as part of a "normal" test run? | 22:44 |
| mtaylor | or do you just want a functests env in tox.ini? | 22:44 |
| *** torgomatic has quit IRC | 22:45 | |
| notmyname | unit tests, definitely. func tests are tricky (they require a running swift install) so they probably should be in a separate env | 22:45 |
| mnewby | notmyname: hi | 22:46 |
| mtaylor | hey! it's mnewby | 22:46 |
| notmyname | mnewby: hi. looking at your tox patch for swift | 22:46 |
| *** torgomatic has joined #openstack-dev | 22:47 | |
| mnewby | So should tox call .unittests instead of nosetests? | 22:48 |
| notmyname | mnewby: I've got some questions (see above about .unittests vs .functests, not to mention .probetests) | 22:48 |
| *** cp16net has quit IRC | 22:48 | |
| mnewby | I'm not sure teh func and probe tests should be executed by tox. | 22:48 |
| mnewby | Tox builds the execution environment, but I'm not sure that's useful for more than unittests. | 22:49 |
| mnewby | func and probe tests require a running swift installation, and I don't think tox is right for that. | 22:49 |
| notmyname | mnewby: I'm leaning towards tox replacing (eventually) our .*tests runners | 22:49 |
| notmyname | ah ok | 22:49 |
| notmyname | so then just .unittests | 22:50 |
| mnewby | mtaylor: Your thoughts? | 22:50 |
| mnewby | I think just unittests, yes. | 22:50 |
| notmyname | which leads to another question. if you aren't doing anything with our test runners, why does your patch change them? | 22:50 |
| mtaylor | mnewby: looking at the code, I think I agree - just unittests. | 22:50 |
| mnewby | I added default nose configuration to setup.cfg | 22:50 |
| *** mfer has quit IRC | 22:50 | |
| mnewby | So the .*tests were updated to remove config that was set in setup.cfg[ | 22:51 |
| notmyname | ah. and nosetests reads that? (I'm not a nose expert) | 22:51 |
| mnewby | (this included config for the nose plugin) | 22:51 |
| mtaylor | yup. | 22:51 |
| mnewby | Yes, nose will read default configuration from setup.cfg. | 22:51 |
| notmyname | ah ok then | 22:51 |
| mnewby | Sorry, meant to say 'included config for the openstack nose plugin' | 22:51 |
| mnewby | .unittests could probably go away. Or call tox. I left it because it seemed useful to be able to run the tests consistently against a system-installed swift rather than one in a tox virtualenv. | 22:52 |
| eglynn | jaypipes: yep, delaying the creds complete-ness check seems to do the trick | 22:54 |
| notmyname | so I've got code in a dev environment (on a VM) installed with `python ./setup.py develop`. does tox and .unittests test the same thing in that case? | 22:54 |
| *** lts has quit IRC | 22:54 | |
| eglynn | jaypipes: for nova image-list at least | 22:54 |
| eglynn | jaypipes: pushed a variant of Stuart's patch to milestone-proposed https://review.openstack.org/4728 | 22:54 |
| eglynn | jaypipes: lets get it approved and thru' Jenkins quickly | 22:55 |
| mnewby | notmyname: No. Tox creates (or updates) it's own virtualenv and runs the tests. .unittests runs the tests against whatever python is active in the environment. | 22:55 |
| mnewby | notmyname: tox and .unittests would only be testing the same thing if you were to manually activate the tox virtualenv prior to invoking .unittests. | 22:56 |
| notmyname | or if the current environment matched the tox virtual env | 22:57 |
| mnewby | Yes | 22:57 |
| notmyname | ok. (that's actually true with most of the dev environments) | 22:58 |
| *** zaitcev has quit IRC | 22:59 | |
| *** mfer has joined #openstack-dev | 22:59 | |
| eglynn | bcwaldon: I need a quick turn-around on https://review.openstack.org/4728 | 23:00 |
| eglynn | bcwaldon: its a variant of Stuart's https://review.openstack.org/#change,4350 patch (already approved) with a fix to hopefully sort out the devstack failure that gating the merge | 23:00 |
| notmyname | mnewby: mtaylor: thanks for taking the time to explain it | 23:00 |
| *** mfer has quit IRC | 23:01 | |
| mnewby | np | 23:01 |
| eglynn | Vek, blamar, blamar__: I'm looking to scare up some glance core reviewers for a quick turn-around on https://review.openstack.org/4728 ... context ^^^ | 23:04 |
| blamar | eglynn: taking a look | 23:04 |
| eglynn | blamar: thanks! | 23:04 |
| *** andrewsmedina has joined #openstack-dev | 23:05 | |
| eglynn | blamar: only diff with previously approved patch is http://www.pastie.org/3492089 | 23:05 |
| eglynn | blamar: avoids premature creds completeness checking on nova image-list | 23:06 |
| *** Gordonz has quit IRC | 23:07 | |
| *** nati_ueno has joined #openstack-dev | 23:07 | |
| *** torgomatic has quit IRC | 23:12 | |
| *** torgomatic has joined #openstack-dev | 23:13 | |
| *** mdomsch has quit IRC | 23:13 | |
| vishy | smoser: ping | 23:14 |
| *** kbringard has quit IRC | 23:20 | |
| blamar | eglynn: haven't forgotten about you, gotta run but going to leave some quick questions on it | 23:27 |
| eglynn | blamar: cool, laters ... | 23:27 |
| *** torgomatic has quit IRC | 23:28 | |
| *** jakedahn has quit IRC | 23:28 | |
| *** torgomatic has joined #openstack-dev | 23:29 | |
| *** flaviamissi has joined #openstack-dev | 23:31 | |
| *** flaviamissi has quit IRC | 23:33 | |
| *** jakedahn__ has joined #openstack-dev | 23:45 | |
| *** mattray has quit IRC | 23:48 | |
| *** vricci has joined #openstack-dev | 23:48 | |
| *** bhall has quit IRC | 23:49 | |
| *** bsza has quit IRC | 23:49 | |
| *** vincentricci has quit IRC | 23:50 | |
| *** vricci is now known as vincentricci | 23:50 | |
| *** _adjohn has joined #openstack-dev | 23:50 | |
| *** adjohn has quit IRC | 23:51 | |
| *** _adjohn is now known as adjohn | 23:51 | |
| *** jakedahn__ has quit IRC | 23:53 | |
| *** jakedahn__ has joined #openstack-dev | 23:53 | |
| *** andrewbogott_afk is now known as andrewbogott | 23:53 | |
| eglynn | blamar: great catch, you're absolutely right about the endpioint['publicURL'] | 23:57 |
| eglynn | blamar: returning the entire dict caused the glance client to fail in a nasty way, fixed up in patch set 2 | 23:57 |
| eglynn | blamar: I'm just trying to get Stuart's patch knocked into shape for E4, so I'm at a bit of disadvantge in terms of explaining how that issue wasn't picked up in testing | 23:57 |
| * eglynn suspects it was falling back to v1 auth in the tests | 23:57 | |
| openstackgerrit | Verification of a change to openstack/melange failed: Add the tenant_id to the block data https://review.openstack.org/4730 | 23:57 |
| *** jakedahn__ is now known as jakedahn | 23:58 | |
| *** milner has joined #openstack-dev | 23:59 | |
| *** jakedahn__ has joined #openstack-dev | 23:59 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!